FO

UN

DA

TIO

N F

OR

DE

FE

NS

E O

F D

EM

OC

RA

CIE

S Islamic State Networks in TurkeyMerve Tahiroglu & Jonathan Schanzer

March 2017

A division of theFOUNDATION FOR DEFENSE OF DEMOCRACIES

Washington, DC

FDD PRESS

Merve Tahiroglu

Jonathan Schanzer

March 2017

Islamic State Networks in Turkey

Islamic State Networks In Turkey

Table of Contents

INTRODUCTION ...............................................................................................................................................1

TURKISH RADICAL ISLAM BEFORE THE ISLAMIC STATE .................................................................8

Turkey’s Jihadist Foreign Fighters Before Syria ...........................................................................................10

TURKEY’S ROLE IN THE SYRIAN JIHAD: 2013-2016 ..............................................................................11

Turkish IS Militants in Syria and Iraq ..........................................................................................................15

The Kurdish Factor ..........................................................................................................................................15

Turkey and IS at War .......................................................................................................................................16

ISLAMIC STATE NETWORKS IN TURKEY.................................................................................................17

ONLINE RADICALIZATION: JIHADISTS ON TURKISH SOCIAL MEDIA ........................................21

The Islamic State’s Turkish-Language Activity on Twitter .........................................................................22

SGA’s Study of IS on Turkish-Language Twitter ..........................................................................................23

SECURITY IN POST-COUP TURKEY: GRIM PROSPECTS .....................................................................24

RECOMMENDATIONS ....................................................................................................................................26

Appendix I - Table I: Islamic State-Linked Attacks in Turkey ........................................................................27

Appendix II - Table II: Anti-Islamic State Operations in Turkey in 2016 ....................................................28

Appendix III - SGA Turkish Social Media Study .............................................................................................33

Islamic State Networks in Turkey

Page 6

IntroductionThe year 2016 was catastrophic for Turkey. At least 30 terror attacks across the country took more than 300 lives.1 Ankara survived a bloody military coup attempt in July, which claimed the lives of an additional 290. In a massive purge that ensued, more than 100,000 civil servants, academics and journalists across the political spectrum were either sacked or detained.2 The economy was downgraded by nearly all of the major credit-rating agencies.3 The military formally joined the Syrian civil war, primarily to carve out a long-desired “safe zone” across the border. And, in a historic moment in December, a Turkish police officer assassinated the Russian ambassador to Ankara. Turkish citizens spent half of the year under a state of emergency, which is still in effect.

The nation rang in 2017 with another devastating terror attack, this time at an iconic Istanbul nightclub, Reina, on New Year’s Eve. The mass shooting killed 39 people, becoming the deadliest attack that the Islamic State (IS) ever claimed in Turkey,4 and the eighth mass assault tied to the group since 2015.5 More than 150 people, many of them tourists, have been killed by IS in Turkey in the last year alone.6

1. Including the suicide bombers, the exact number of deaths for 2016 is 325, based on a tally of the figures listed in “Bir buçuk yılda 33 bombalı saldırıda 461 kişi hayatını kaybetti; 363’ü sivil (461 people, 363 civilians lost their lives in 33 bombings in a year and a half ),” Diken (Turkey), December 12, 2016. (http://www.diken.com.tr/bir-bucuk-yilda-33-bombali-saldirida-461-kisi-hayatini-kaybetti-363u-sivil/)2. “Purge in Numbers,” Turkey Purge, February 2, 2017. (http://turkeypurge.com/purge-in-numbers)3. S&P and Moody’s downgraded Turkey’s ratings in 2016, while Fitch joined them on January 27, 2017. Mehreen Khan, “Turkey cut to junk by Fitch, losing last major investment-grade rating,” Financial Times (UK), January 27, 2017. (https://www.ft.com/content/feb883bd-577f-31c6-b977-1c8082eccbad)4. Thomas Joscelyn, “Islamic State Claims Responsibility for New Year’s Day Attack at Istanbul Nightclub,” FDD’s Long War Journal, January 2, 2017. (http://www.longwarjournal.org/archives/2017/01/islamic-state-claims-responsibility-for-new-years-day-attack-at-istanbul-nightclub.php)5. The bombing at a Diyarbakir police station on November 4, 2016, which killed 11 people, was claimed by both IS and the Kurdistan Freedom Falcons (TAK), a Kurdish terror group in Turkey. Based on the target and location of the attack, the author ascribes the attack to TAK. See: Mahmut Bozarslan, “One bomb, three suspects: Who was behind latest Diyarbakir attack?” Al Monitor, November 13, 2016. (http://www.al-monitor.com/pulse/originals/2016/11/turkey-diyarbakir-bombing-who-did-it-isis-pkk.html)6. See Appendix I. 7. “Numan Kurtulmuş’tan Sosyal Medya Uyarısı (Social Media Warning from Numan Kurtulmuş),” CNN Türk (Turkey), January 2, 2017. (http://www.cnnturk.com/video/turkiye/numan-kurtulmustan-sosyal-medya-uyarisi)8. Daren Butler, “Death toll in Istanbul bombings rises to 44: health minister,” Reuters, December 12, 2016. (http://www.reuters.com/article/us-turkey-blast-toll-idUSKBN1410VO); “Kayseri suicide bomber arrived from Kobane in Syria before attack,” Hurriyet Daily News (Turkey), December 19, 2016. (http://www.hürriyetdailynews.com/kayseri-suicide-bomber-arrived-from-kobane-in-syria-before-attack-.aspx?PageID=238&NID=107471&NewsCatID=341)

Following the Reina massacre, the parliament extended the state of emergency for another three months, the second extension since the July 15 coup attempt. Authorities arrested dozens of people and issued an immediate media ban – as they have after every crisis in Turkey, including the assassination of the Russian ambassador in December. Ankara also announced “all social media accounts are being monitored.”7 But the ruling Justice and Development Party (AKP)’s censorship of Turkish media and state of emergency measures have clearly failed to make Turkey safer. The widespread purges of the Turkish military and law enforcement officials have not helped.

To be sure, the Islamic State is just one of the groups that has targeted Turkey in the last two years. In December alone, Kurdish militants conducted three suicide attacks – twin bombings in Istanbul and another one in central Turkey – killing a combined 58 and wounding more than 150.8

The rise of renewed Kurdish radicalism was sadly predictable. In July 2015, Ankara’s peace talks with the separatist Kurdistan Workers’ Party (PKK), a designated terror organization, ended after a two-year ceasefire.

Islamic State Networks in Turkey

Page 7

The PKK has fought the Turkish state for four decades, but no Turkish government had ever negotiated with the group’s imprisoned leader, Abdullah Ocalan, before the AKP’s reign. The 2013 ceasefire had produced a period of unprecedented calm and socio-economic opportunity. Many younger-generation Kurds were infuriated by Turkish President Recep Tayyip Erdogan’s 2015 change of heart vis-à-vis the pursuit of a peaceful resolution of the country’s Kurdish conflict.

Since the so-called “solution process” ended in 2015, PKK-affiliated Kurds have carried their traditional guerilla warfare in Turkey’s southeast from the villages into the cities, and have detonated suicide vests in major Turkish cities including Ankara and Istanbul. Together with its offshoots, such as the Kurdistan Freedom Falcons (TAK),9 the PKK has claimed more than 300 civilian lives in over 30 bombings since July 2015. Included among their targets was the leader of Turkey’s main opposition Republican People’s Party (CHP) Kemal Kilicdaroglu, who survived the attack on his convoy in August but has received more threats since.10

Between the PKK and its more extreme affiliates in Turkey, Ankara has its hands full. But as both fights drag on, it seems increasingly clear that Turkey’s fight against Kurdish militants is steadily undermining its struggle against the Islamic State.

Since the beginning of the Syrian civil war, Turkey’s indifference towards and even tacit support for IS and other jihadists battling the Kurds across its borders

9. For a quick overview of TAK, see Aykan Erdemir, “PKK Offshoot Claims Ankara Attack,” Foundation for Defense of Democracies, February 19, 2016. (http://www.defenddemocracy.org/media-hit/dr-aykan-erdemir-pkk-offshoot-claims-ankara-attack/)10. “Turkish main opposition CHP leader survives PKK attack on motorcade,” Hurriyet Daily News (Turkey), August 25, 2016. (http://www.hurriyetdailynews.com/turkish-main-opposition-chp-leader-survives-pkk-attack-on-motorcade.aspx?PageID=238&NID=103217&NewsCatID=341)

has alarmed its Western allies, particularly the United States. Knowingly or not, Turkey allowed IS and other jihadist groups to establish their cells in Istanbul, Ankara, and other Turkish cities near the Syrian frontier. Turkey’s own radical Islamists have proved easy for IS and other Syria-based jihadists from groups – such as Hay’at Tahrir al-Sham (HTS, also formerly known as Jabhat Fatah al-Sham and Jabhat al-Nusra) and Ahrar al-Sham – to recruit.

Beyond their physical networks in Turkey, the jihadists’ online presence in Turkish is growing. Indeed, social media has become the top recruiting platform for IS and other tech-savvy extremist groups. And while Turkey has imposed draconian media laws, Ankara’s online crackdown on jihadists remains relatively meek, with the AKP showing far greater alarm over anti-government expressions of political dissent.

Until 2015, militants had been preoccupied with the jihad against the Bashar al-Assad regime (and all minority groups) in Syria, sparing the Turks for the large part. Mounting attacks over the last two years, however, make clear that is no longer the case. As Turkey’s territorial designs in northern Syria increasingly clash with those of IS, and the two sides engage in direct military combat with greater intensity, the Islamic State is increasingly inclined to punish the Turks at home. With residual IS networks now spread throughout the country, the worst for Turkey may be yet to come.

Ankara claims that it foiled nearly 350 terror plots last year. If true, that would be an outstanding achievement. But the amount of terror-related bloodshed Turkey has suffered in the last two years is jarring. The prospect of continued violence threatens the country’s security, as well as the stability of its neighbors and allies.

“ As both fights drag on, it seems increasingly clear that Turkey’s fight against Kurdish militants is steadily undermining its struggle against the Islamic State.”

Islamic State Networks in Turkey

Page 8

Turkish Radical Islam Before the Islamic State

Since the founding of the modern Turkish republic, Turks have largely been spared the sort of violent religious zeal found elsewhere in the region. A majority of Turkish citizens follow the Hanafi school of Islam, with the remaining following Shafi, Jafari, and Alevi traditions, all of which have historically opposed the purist religious doctrine of the Salafists.11 Most Islamist Turkish citizens are, therefore, unaccustomed to ultra-conservative Salafi interpretations of the Quran.

But neither Turkey’s secular tradition nor its predominant, moderate Islamic theology has made Turkish citizens immune to radicalization. Modern Turkish history has produced several violent Islamist organizations, appealing to the republic’s Turks and Kurds alike.

The five most notable Islamist extremist groups in Turkey today are Kurdish Hizbullah, Kaplancilar, IBDA-C, Hizb ut-Tahrir, and Tahsiyeciler. At least two of these organizations have direct links to IS, while the others advocate Salafi views in line with the Islamic State’s ideology and contribute to the group’s recruitment efforts. And while some of these groups may now be defunct, they are believed to have provided an infrastructure for IS in Turkey.

11. Selim Koru, “Turkey’s 200-Year War against ‘ISIS,’” The National Interest, July 24, 2015. (http://nationalinterest.org/feature/turkeys-200-year-war-against-isis-13412?page=show)12. For more on this group and its recent activities in Turkey, see Aykan Erdemir and Merve Tahiroglu, “The ISIS attack and Turkey’s Islamist Kurds,” NOW Lebanon, July 30, 2015. (https://now.mmedia.me/lb/en/commentary/565659-the-isis-attack-and-turkeys-islamist-kurds)13. Kadri Gürsel, “New ‘Party of God’ Will Divide Kurdish, Turkish Islamists,” Al Monitor, December 23, 2012. (http://www.al-monitor.com/pulse/originals/2012/al-monitor/hizbullah-turkey-islamist.html)14. “The Kurdish Hizbullah in Turkey,” Oxford Islamic Studies, accessed February 2, 2017. (http://www.oxfordislamicstudies.com/article/opr/t236/e0317#e0317-s3)15. Between the 2014 and 2015 elections, Huda-Par competed in a total of 13 provinces. The author compared the number of votes won by Huda-Par candidates in each election by province, as published in the Turkish parliament’s official website. To calculate the party’s maximum electoral capacity, the author tallied the highest number of votes won in each of the 13 provinces, reaching a total of 100,533 votes. 16. Karışık Video İzle, “Halis Bayancuk’tan IŞİD’e Tam Destek (Full Support for ISIL from Halis Bayancuk),” YouTube, December 21, 2014. (https://www.youtube.com/watch?v=-pWUp8xohhg)17. “İstanbul’da bayram namazı kılan IŞİD’den Türkiye’ye tehdit: Aklınızı başınıza alın (ISIL, Holding Eid Prayers in Istanbul, Threatens Turkey” Cumhuriyet (Turkey), July 18, 2015. (http://www.cumhuriyet.com.tr/haber/turkiye/325675/istanbul_da_bayram_namazi_kilan_ISiD_den_Turkiye_ye_tehdit__Aklinizi_basiniza_alin.html)

Kurdish Hizbullah / Huda-Par

Hizbullah, a Sunni Kurdish organization unrelated to its Lebanese namesake, is among the deadliest Islamist-extremist terror groups in Turkish history, and it continues to enjoy widespread support among Turkey’s Islamist Kurds today.12 Inspired by Iran’s Islamic revolution in 1979, Hizbullah has endeavored to create an Islamic state for Turkey’s Kurds since the 1980s.13 The violence it unleashed against secular-nationalist Kurds left hundreds dead during the intra-Kurdish conflict in the 1990s.

In December 2012, Hizbullah rebranded as Huda-Par, a legal political party. But the new outfit is far from peaceful. Since the IS-Kurdish fighting in northern Syria intensified in 2014, Huda-Par supporters have repeatedly clashed with secular and nationalist Kurds, who argue that Huda-Par supports the Islamic State.14 Based on Huda-Par’s performance in the March 2014 municipal and the June 2015 general elections, the group today enjoys the support of an estimated 100,000 Kurds.15

Halis Bayancuk, also known by his nom de guerre Abu Hanzala, is a Turkish Islamist notorious for his pro-IS lectures and public statements. He is the son of a former Hizbullah convict, and has advocated for both the Islamic State and Hizbullah on social media.16 In both 2014 and 2015, Bayancuk led prayers at an annual Eid picnic in an Istanbul suburb, reportedly organized by IS sympathizers.17

Islamic State Networks in Turkey

Page 9

“Kaplancilar” / Anatolian Federated Islamic State (Caliphate State or Kalifasstaat)

Kaplancilar, also known as the Anatolian Federated Islamic State and “Kalifasstaat,” is a Germany-based group founded in 1984 by Cemalettin Kaplan, a Turkish asylum seeker sentenced to death in Turkey. The movement’s goal is to create an Islamic state in Turkey, modeled after the Islamic Republic of Iran.18 After Kaplan’s death in 1994, his son, Metin Kaplan, took over the leadership and has attracted 1,100 followers across Germany.19

According to Turkish security forces, the group planned two attacks inside Turkey in October 1998 – including a plot to bomb the mausoleum of Turkey’s founding father, Mustafa Kemal Ataturk, on the republic’s 75th anniversary. Turkish authorities arrested 23 members connected to these plots. Another 40 were arrested in Germany that year at a demonstration for the release of Kaplan from jail, which some 500 members attended.20

The group was banned in Germany after the September 11, 2001 attacks in the U.S. – becoming the first organization to be banned under revised German anti-terror laws.21 German authorities revoked Kaplan’s refugee status and extradited him to Turkey on terrorism charges. Kaplan was sentenced to life imprisonment in Turkey. His organization remains intact, but has been rather dormant.

18. Brian R. Farmer, Radical Islam in the West: Ideology and Challenge, (Jefferson, NC: McFarland & Co., 2011), page 43.19. “Profile: The Caliph of Cologne,” BBC (UK), May 27, 2004. (http://news.bbc.co.uk/2/hi/europe/1705886.stm)20. Brian R. Farmer, Radical Islam in the West: Ideology and Challenge, (Jefferson, NC: McFarland & Co., 2011), page 43.21. “‘Caliph of Cologne’ Loses Battle,” BBC (UK), May 26, 2004. (http://news.bbc.co.uk/2/hi/europe/3751341.stm)22. Ely Karmon, “Islamic terrorist activities in turkey in the 1990s,” Terrorism and Political Violence, December 1998, page 107. (http://dx.doi.org/10.1080/09546559808427484)23. Ibid, pages 107-109. 24. Amiram Barkat, Yoav Stern and Yossi Melman, “Attacks on Istanbul Synagogues Kill 24,” Haaretz (Israel), November 16, 2003. (http://www.haaretz.com/attacks-on-istanbul-synagogues-kill-24-1.105755)25. Ely Karmon, “The Synagogue Bombings in Istanbul: Al-Qaeda’s New Front?” The Washington Institute for Near East Policy, November 18, 2003. (http://www.washingtoninstitute.org/policy-analysis/view/the-synagogue-bombings-in-istanbul-al-qaedas-new-front) 26. Mustafa Akyol, “Another Turkish witch hunt begins,” Al Monitor, December 15, 2014. (http://www.al-monitor.com/pulse/originals/2014/12/turkey-gulen-movement-media-freedom-crackdown.html)

Great Eastern Islamic Raiders’ Front (IBDA-C)

Islami Buyuk Dogu Akincilari Cephesi (IBDA-C), or the Great Eastern Islamic Raiders’ Front, is a Sunni Islamist movement active in Turkey since the mid-1970s. The group’s declared mission is to overthrow the secular Turkish government and establish a caliphate. IBDA-C became increasingly violent in the 1990s, threatening and targeting Turkey’s artists, public intellectuals, and Jewish and Christian charities. In the year 1994 alone, the group may have been responsible for up to 90 terror attacks, including five bombings.22

IBDA-C is particularly infamous for its anti-Semitism. After issuing death threats against Professor Yuda Yurum, the leader of Ankara’s Jewish community in 1995, the group’s members attempted to assassinate Yurum by placing a bomb in his car that June.23

The organization also claimed responsibility for the twin synagogue bombings in Istanbul in November 2003, which killed a total of 24 people and injured hundreds.24 While those attacks are attributed to al-Qaeda and were conducted with the global jihadist network’s assistance, at least one of the perpetrators was identified as a former IBDA-C member.25

Tahsiyeciler

Led by the Islamist journalist Mehmet Dogan, Tahsiyeciler is a small group sympathetic to al-Qaeda, though with no known history of violence in Turkey.26 On January 22, 2010, Turkish police carried out

Islamic State Networks in Turkey

Page 10

simultaneous raids against Tahsiyeciler in 16 Turkish cities, detaining 132 and arresting 38 suspects.27 Authorities reportedly found handguns, hunting rifles, grenades, a smoke bomb, electronic parts for explosives, and a large cache of ammunition.28 They also found a video, broadcast by Turkish media, in which Dogan calls on his followers to build bombs and mortars in their homes and to join the jihad.29

Along with Dogan, Mustafa Kaplan – a former columnist for the Islamist daily Vakit – and Mehmet Nuri Turan are identified as key members.30 Mustafa Gunes, recently identified as the top recruiter for the Islamic State in the central Anatolian city of Konya, reportedly had ties to Kaplan before forming his own al-Qaeda network in Turkey in 2008.31

Most experts do not consider Tahsiyeciler part of the global al-Qaeda network, but rather as the victim of a politically motivated witch-hunt.32 Still, it cannot be ignored that the group continues to profess extremist views and spread pro-jihadist propaganda. In 2014, Dogan proclaimed on CNN Turk that he loved Osama bin Laden.33

27. “Askeri Istihbarat da Tahşiyeciler’i Takibe Almış (Military Intelligence Was Also Monitoring Tahşiyeciler),” T24 (Turkey), January 3, 2015. (https://t24.com.tr/haber/askeri-istihbarat-da-tahşiyecileri-takibe-almis,282442)28. “General Staff Intelligence Warned Military Units About Pro-al-Qaeda Tahşiyeciler,” Facts on Turkey, January 3, 2015. (http://factsonturkey.org/14972/)29. Guyverland, “CNN Türk Tahşiye Senaryosunda Gizlenen Gerçekler (CNN Türk Hidden Secrets from Tahsiye Scenario),” YouTube, December 16, 2014. (https://www.youtube.com/watch?v=eNnXlRUfK0Q&ab_channel=guyverland) 30. “General Staff Intelligence Warned Military Units About Pro-al-Qaeda Tahşiyeciler,” Facts on Turkey, January 3, 2015. (http://factsonturkey.org/14972/)31. Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)32. Mustafa Akyol, “WikiLeaks sheds light on Turkey’s mysterious ‘Annotators,’” Al Monitor, December 22, 2014. (http://www.al-monitor.com/pulse/originals/2014/12/turkey-wikileaks-mysterious-annotators-al-qeade.html)33. Leonard Brojan, “Mehmet Doğan: Ladin’i Severim (Mehmet Doğan: I Love Ladin),” YouTube, December 27, 2014. (https://www.youtube.com/watch?v=scuI8e2Dqpk&ab_channel=leonardbrojan)34. “Hizb-ut Tahrir: Terör örgütü Değiliz, Silah Kullanımına Karşıyız (Hizb-ut Tahrir: We Are Not a Terror Organization, We Oppose Using Weapons),” T24 (Turkey), September 28, 2014. (http://t24.com.tr/haber/hizb-ut-tahrir-teror-orgutu-degiliz-silah-kullanimina-karsiyiz,272151)35. Hizb ut-Tahrir website, accessed February 2, 2017. (http://hizb-turkiye.org/)36. “Radical Islamist Hizb ut-Tahrir calls for caliphate in grand meeting in Ankara,” Hurriyet Daily News (Turkey), March 8, 2016. (http://www.hürriyetdailynews.com/radical-islamist-hizb-ut-tahrir-calls-for-caliphate-in-grand-meeting-in-ankara.aspx?pageID=238&nID=96179&NewsCatID=341)

Hizb ut-Tahrir

Hizb ut-Tahrir is a global radical Islamist organization, reportedly with branches in over 50 countries. The group was founded in Jerusalem in 1953, and seeks to reestablish a caliphate for the Muslim world. Its Turkish branch has been operational since the 1960s, and has been outlawed by the Turkish government.34 Hizb ut-Tahrir has an active Turkish-language website35 and magazine, Koklu Degisim, which means “radical change.” Most recently in March 2016, the group held two conferences in Istanbul and Ankara to mourn the 92nd anniversary of the caliphate’s abolishment. The “International Caliphate Conference” in Istanbul attracted around 5,000 supporters; the one in Ankara, themed “Caliphate: A Dream or Soon To-Be Reality,” also amassed thousands.36

Turkey’s Jihadist Foreign Fighters Before Syria

Extremist Turks have also historically been animated by global conflicts concerning Muslim populations. Turkish jihadists have been found among the ranks of foreign fighters in several recent conflict zones abroad.

Islamic State Networks in Turkey

Page 11

Most notably, both the Afghan war in the 1980s and the Bosnian war in the 1990s drew jihadist Turks to their territories.37 At least five Turkish citizens were reportedly killed in the Afghan civil war, and 21 in the Bosnian conflict. Turkish foreign fighters have even been documented in conflicts farther from their homeland, such as in Ogaden, Ethiopia; Kashmir, India; and Chechnya.38

The number of Turkish jihadists soared with the advent of al-Qaeda. In the early 2000s, the number of Turkish foreign fighters in Iraq and Afghanistan reportedly reached hundreds.39 Turkish al-Qaeda members have reportedly carried out suicide attacks against U.S. forces in Iraq and now figure prominently among the group’s list of “martyrs.”40 Several others assumed leadership positions in al-Qaeda, including Serdal Erbasi, who was identified as al-Qaeda’s regional commander for Turkey and the Caucuses in 2010.41 Reports in 2010 pointed to a proliferation of Turkish-language jihadi websites, as well as the growing number of Turkish fighters in the Afghanistan-Pakistan jihadist theater. Analysts warned that Turkey had turned into a “gateway” for al-Qaeda’s funds and recruits.42

37. Serhat Erkman, “Suriye ve Irak’ta Savaşan Türkiyeli Mücahitler (Turkish Foreign Fighters in Syria and Iraq),” Al Jazeera Turk (Turkey), May 25, 2015. (http://www.aljazeera.com.tr/gorus/suriye-ve-irakta-savasan-turkiyeli-mucahitler)38. “Turkey’s Jihadists,” Turkey Wonk: Nuclear and Political Musings in Turkey and Beyond, April 28, 2015. (https://turkeywonk.wordpress.com/2015/04/28/turkeys-jihadists/); “Turkish Foreign Fighters in Ogaden,” The Loopcast, January 26, 2015. (http://www.theloopcast.com/e/turkey-foreign-fighters-in-ogaden/)39. Serhat Erkman, “Suriye ve Irak’ta Savaşan Türkiyeli Mücahitler (Turkish Foreign Fighters in Syria and Iraq),” Al Jazeera Turk (Turkey), May 25, 2015. (http://www.aljazeera.com.tr/gorus/suriye-ve-irakta-savasan-turkiyeli-mucahitler)40. Mehmet Orhan, “Al-Qaeda: Analysis of the Emergence, Radicalism, and Violence of a Jihadist Action Group in Turkey,” Turkish Studies, 2010. (http://dx.doi.org/10.1080/14683849.2010.483846)41. “El Kaide komutanı ‘Ebuzer’ Türk çıktı (Al Qaeda Commander ‘Ebuzer’ Turns Out To Be A Turk),” Milliyet (Turkey), January 20, 2010. (http://www.milliyet.com.tr/el-kaide-komutani--ebuzer--turk-cikti-gundem-1188281/)42. Simon Cameron-Moore, “Analysis: Al Qaeda fishes for Turks seeking jihad,” Reuters, October 27, 2010. (http://www.reuters.com/article/us-turkey-security-idUSTRE69Q3EN20101027)43. “Turkey’s Jihadists,” Turkey Wonk: Nuclear and Political Musings in Turkey and Beyond, April 28, 2015. (https://turkeywonk.wordpress.com/2015/04/28/turkeys-jihadists/)44. Aaron Stein, “The Islamic State in Turkey: A Deep Dive Into a Dark Place,” War On the Rocks, April 6, 2016. (http://warontherocks.com/2016/04/the-islamic-state-in-turkey-a-deep-dive-into-a-dark-place/)45. Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)46. “Erdoğan To Assad: Stop the Violence, Implement Reforms,” The Jerusalem Post, June 14, 2011. (http://www.jpost.com/Middle-East/Erdoğan-to-Assad-Stop-the-violence-implement-reforms)

Returning Turkish jihadist foreign fighters have always posed a security threat to Turkey. Jihadists who returned from the Afghan war of the 1980s came to establish al-Qaeda’s Turkish cells. Perpetrators of the 2003 al-Qaeda attacks in Istanbul were returnees. Several Turkish jihadists reportedly killed in the Syrian civil war since 2011 are believed to have previously fought in the Afghan and Bosnian wars.43 Today, those returnees appear to have formed the backbone of the IS network in the country.44 Most, if not all, key Turkey-based IS recruiters today are believed to have links to the Afghan jihads of the 1980s and 2001.45

Turkey’s Role in the Syrian Jihad: 2013-2016

When the uprisings against Syrian President Bashar al-Assad began in 2011, few observers foresaw the country’s drift into civil war. That June, Erdogan called on his Syrian counterpart to halt the violence and implement reforms.46 By August, it was clear that Assad had no interest in either. U.S. President Barack Obama called on the Syrian president to resign that

Islamic State Networks in Turkey

Page 12

month,47 and Erdogan followed suit in November.48 The Assad regime, however, proved resilient. The Syrian opposition soon splintered into armed factions with competing interests, and the country descended into civil war.

Turkey threw its full support behind the opposition – harboring and mobilizing dissidents, arming rebels, and allowing free passage along the Syrian-Turkish border. Even as the war raged along Turkey’s longest and most porous frontier, Ankara refused to close its border, citing humanitarian concern for refugees. But jihadists began to exploit the security vacuum, including Jabhat al-Nusra (later JFS and now HTS), the Syrian affiliate of al-Qaeda, right across the border. One former U.S. official claims that Ankara was deliberately helping the jihadists and acknowledged it.49 While the U.S. designated al-Nusra as a terror organization in December 2012, Ankara publicly opposed the move for months,50 only finally designating the group in June 2014.51 The damage, however, was already done. Thousands of Turkish jihadists reportedly traveled to Syria to join extremist groups between 2013 and late 2014, some under the pretense of doing humanitarian

47. Scott Wilson and Joby Warrick, “Assad must go, Obama says,” The Washington Post, August 18, 2011. (https://www.washingtonpost.com/politics/assad-must-go-obama-says/2011/08/18/gIQAelheOJ_story.html?utm_term=.aa3df29f7a07)48. Şebnem Arsu, “Turkish Premier Urges Assad to Quit in Syria,” The New York Times, November 22, 2011. (http://www.nytimes.com/2011/11/23/world/middleeast/turkish-leader-says-syrian-president-should-quit.html)49. Hannah Allam, “Once a top booster, ex-U.S. envoy no longer backs arming Syrian rebels,” McClatchy, February 18, 2015. (http://www.mcclatchydc.com/news/nation-world/world/article24780202.html)50. Semih İdiz, “Pressure Mounts on Turkey Over Radical Groups in Syria,” Al Monitor, October 15, 2013. (http://www.al-monitor.com/pulse/originals/2013/10/turkey-must-control-jihadists-entering-syria.html) 51. Dorian Jones, “Turkey Designates Al-Nusra Front as a Terrorist Organization,” Voice of America, June 4, 2014. (http://www.voanews.com/a/turkey-designates-al-nusra-front-as-a-terrorist-organization/1929675.html)52. “Turkey’s Jihadists,” Turkey Wonk: Nuclear and Political Musings in Turkey and Beyond, April 28, 2015. (https://turkeywonk.wordpress.com/2015/04/28/turkeys-jihadists/); “Turkish police detain 28 in anti-al-Qaeda op, raid on İHH office,” Hurriyet Daily News (Turkey), January 14, 2014. (http://www.hürriyetdailynews.com/turkish-police-detain-25-at-ihh-office-in-anti-al-qaeda-raid.aspx?pageID=238&nID=61000&NewsCatID=341)53. Eric Schmitt, “C.I.A. Said to Aid in Steering Arms to Syrian Opposition,” The New York Times, June 21, 2012. (http://www.nytimes.com/2012/06/21/world/middleeast/cia-said-to-aid-in-steering-arms-to-syrian-rebels.html)

work. At least one future IS member, Yakup Aktulum, said he first traveled to Syria in 2013 with the Humanitarian Aid Foundation (IHH) – an Islamist Turkish charity with suspected ties to al-Qaeda.52

In 2012, the Arab Spring had energized the Muslim Brotherhood and its affiliates in the region – presenting Turkey a rare opportunity to exert its influence in the domestic politics of its neighbors. As Islamists took power in Egypt and Tunisia, and a region-wide shift toward political Islam appeared on its way, Turkey’s own Islamists – having won consecutive elections for a decade – found themselves uniquely positioned to lead the transformation and extend their political influence in the neighborhood. In no country could such influence be more rewarding for Turkey than in its immediate neighbor Syria. Ankara displayed an ideological bias in favor of the Syrian opposition’s Islamist factions, even as those groups rapidly turned violent and radical, as a natural consequence of the war.

That ideological bias eventually came to undermine Turkish-U.S. plans to provide joint covert support for the Syrian opposition. While the Turkish intelligence coordinated the transfer of Saudi and Qatari arms to various rebel groups in northern Syria, the CIA sent officers to southern Turkey to help vet the Syrian opposition receiving the aid.53 But vetting became a source of disagreement for the two NATO allies. By October 2012, U.S. officials conceded that the arms

“ One former U.S. official claims that Ankara was deliberately helping the jihadists and acknowledged it.”

Islamic State Networks in Turkey

Page 13

flow from Turkey was mostly helping the jihadists.54 By the following October, Washington was openly pointing the finger to Turkey as the reason why.55

Erdogan and his then-Foreign Minister Ahmet Davutoglu were determined to follow through with their mission to arm the rebels of their choice, with or without U.S. support. But that grew complicated on January 1, 2014, when the governor of Hatay, a Turkish town bordering Syria’s Idlib province west of Aleppo, stopped a truck full of weapons headed to Syria. Turkish press reported that the truck driver had told the governor he was on a mission on behalf of Turkey’s National Intelligence Agency (MIT).56 The government later stated that the truck carried humanitarian aid for Syria’s Turkmen. But original reports indicated that the trucks were headed to Kilis, closer to jihadist-held territories near Syria’s Azaz.57

Later that month, a similar incident was reported in Adana. On January 19, Turkish gendarmerie stopped a truck full of weapons, including artillery and rockets, which turned out to belong to the MIT.58 When the gendarmerie called a senior prosecutor to the scene, as standard protocol, and blocked the trucks from moving per the prosecutor’s orders, things escalated. The governor ordered the police to withdraw from the

54. David E. Sanger, “Rebel Arms Flow Is Said to Benefit Jihadists in Syria,” The New York Times, October 14, 2012. (http://www.nytimes.com/2012/10/15/world/middleeast/jihadists-receiving-most-arms-sent-to-syrian-rebels.html)55. Adam Entous and Joe Parkinson, “Turkey’s Spymaster Plots Own Course on Syria,” The Wall Street Journal, October 10, 2013. (http://www.wsj.com/articles/SB10001424052702303643304579107373585228330)56. “Intel not tasked with carrying weapons, Turkish main opposition leader says,” Hurriyet Daily News (Turkey), January 20, 2014. (http://www.hürriyetdailynews.com/intel-not-tasked-with-carrying-weapons-turkish-main-opposition-leader-says.aspx?pageID=238&nID=61279&NewsCatID=338)57. “Hatay’da mühimmat yüklü TIR şoku (Weapon-Loaded Truck Shock in Hatay),” CNN Türk (Turkey), January 1, 2014. (http://www.cnnturk.com/turkiye/hatayda-muhimmat-yuklu-tir-soku)58. Metin Gurcan, “MIT gun truck bust lands Turkish gendarmes in prison,” Al Monitor, April 20, 2015. (http://www.al-monitor.com/pulse/originals/2015/04/turkey-scapegoating-upsets-gendarmerie-shipping-weapons.html)59. Ibid.60. Holly Yan and Samira Said, “Floodgates open as ISIS bridges victories between Syria and Iraq,” CNN, July 9, 2014. (http://www.cnn.com/2014/07/08/world/meast/syria-civil-war/); Liz Sly and Ahmed Ramadan, “Insurgents seize Iraqi city of Mosul as security forces flee,” The Washington Post, June 10, 2014. (https://www.washingtonpost.com/world/insurgents-seize-iraqi-city-of-mosul-as-troops-flee/2014/06/10/21061e87-8fcd-4ed3-bc94-0e309af0a674_story.html)61. Zeke J. Miller, “Full Text of Obama’s Address on ISIS,” TIME, September 10, 2014. (http://time.com/3320666/obama-isis-speech-full-text/)62. “U.S.: More than 40 nations joined coalition against ISIS,” Al Arabiya (UAE), September 8, 2014. (http://english.alarabiya.net/en/News/world/2014/09/08/US-More-than-40-nations-joined-coalition-against-ISIS.html)

scene, against the prosecutor’s wishes. The incident made clear that there was no room for the military or law enforcement to intervene in what appeared to be state policy. Erdogan dismissed the incident as a secret plot to overthrow his government, despite clear evidence that the gendarmes merely did their jobs: They identified a suspicious truck, searched it, and tried to prevent it from crossing into Syria.59

The rise of the Islamic State in the first half of 2014 dramatically altered the threat matrix in Syria. IS was fiercer than other jihadist groups, such as al-Qaeda, and demonstrated an unparalleled ability to conquer, hold, and govern territory. By July 2014, IS controlled the Iraqi cities of Fallujah and Mosul, as well as Syrian cities between Deir Ezzor and the Iraqi border.60 In September, the United States formally declared war61 and formed a coalition with over 40 countries to fight IS.62

Ankara, however, did not share its allies’ enthusiasm. It was slow to take military action against the group, or crack down on the smuggling of militants, oil, and other goods across the border until the summer of 2015. Turkey did not even allow the anti-IS coalition to operate from its NATO airbase in Incirlik near IS territory in Syria. For a year, coalition airstrikes had to

Islamic State Networks in Turkey

Page 14

be carried out from bases farther away in Cyprus, the United Arab Emirates, and Qatar.63

Admittedly, Turkey likely worried about blowback in response to potential Turkish military action. By March 2014, IS members had already demonstrated their ability to operate in Turkey by carrying out a small-scale attack in central Anatolia, shooting two security officials and a bystander.64 The group had also taken Turkish diplomatic staff hostage in Mosul, and threatened a small Turkish outpost in Syria – the Ottoman tomb of Suleiman Shah, then guarded by 38 Turkish soldiers.65 IS also controlled two towns directly across the Turkish border, Jarabulus and Tel Abyad, by mid-2014. Ankara likely did not want to provoke the group, given its proximity.

Ankara’s unwillingness to engage IS militarily is understandable. But its indifference towards the group’s activities within and across Turkish territory was irresponsible. Remarkably, Turkey took few steps to seal its Syrian border, despite repeated entreaties from Washington, even after IS captured bordering towns. While the crossings between Turkey and IS territory were officially shut in 2014, smuggling through the frontier continued apace.66

Materials, such as ammonium nitrate fertilizers, flowed into then-IS strongholds in northern Syria from Turkey. Fertilizers could be used for agricultural purposes, but

63. John Cappello, Patrick Megahan, John Hannah, and Jonathan Schanzer, “Covering the Bases: Reassessing U.S. Military Deployments in Turkey After the July 25 Coup D’état,” Foundation for Defense of Democracies, August 2016. (http://www.defenddemocracy.org/content/uploads/documents/Covering_the_Bases.pdf ) 64. Semih İdiz, “ISIS emerges as threat to Turkey,” Al Monitor, March 25, 2014. (http://www.al-monitor.com/pulse/originals/2014/03/isis-threat-turkey-syria-jihadists.html)65. “Turkey vows to respond to any attack on Süleyman Shah Tomb,” Hurriyet Daily News (Turkey), March 23, 2014. (http://www.hürriyetdailynews.com/turkey-vows-to-respond-to-any-attack-on-suleyman-shah-tomb.aspx?pageID=238&nID=63967&NewsCatID=359)66. Tim Arango and Eric Schmidt, “A Path to ISIS, Through a Porous Turkish Border,” The New York Times, March 9, 2015. (https://www.nytimes.com/2015/03/10/world/europe/despite-crackdown-path-to-join-isis-often-winds-through-porous-turkish-border.html) 67. Tolga Tanis, “DEAŞ’ın Türk malı potasyum nitratları (ISIL’s Turkish-Made Potassium Nitrates),” Hurriyet Daily News (Turkey), December 18, 2016. (http://www.hürriyet.com.tr/yazarlar/tolga-tanis/deasin-turk-mali-potasyum-nitratlari-40310131)68. “Hürriyet gazetesi Tolga Tanış’ı görevden aldı (Daily Hürriyet Removes Tolga Tanis from His Post,” Cumhuriyet (Turkey), January 10, 2017. (http://www.cumhuriyet.com.tr/haber/turkiye/658236/Hürriyet_gazetesi_Tolga_Tanis_i_gorevden_aldi.html)69. Ben Hubbard and Karam Shoumali, “Fertilizer, Also Suited for Bombs, Flows to ISIS Territory From Turkey,” The New York Times, May 4, 2015. (http://www.nytimes.com/2015/05/05/world/europe/fertilizer-also-suited-for-bombs-flows-to-isis-territory-from-turkey.html)70. Ibid.

ammonium nitrate is also a common ingredient for making explosives. Investigating the flow of dual-use goods from Turkey into the Islamic State, Turkish journalist Tolga Tanis noted the negligible export of ammonium nitrate to Syria before the civil war (only twice, in 2003 and then in 2008) as compared to the export of more than 13,000 tons between 2013 until the Turkish ban in 2016.67 The Turkish government was enraged by the expose, and put pressure on Hurriyet, which removed Tanis from his post as the paper’s Washington correspondent three weeks later.68

In May 2015, The New York Times reported that despite the border closure, large quantities of fertilizers were still being permitted to cross the Turkish border gate at Akcakale, Sanliurfa into IS-controlled Tel Abyad.69 IS documents captured by Syrian Kurdish forces who liberated Tel Abyad that fall proved that the jihadists were indeed using the Turkish fertilizers for manufacturing bombs. In one instance, a document showed that IS bought up to 125 tons of fertilizer from the Turkish side.70

IS also uses Turkish potassium nitrate to make improvised explosive devices (IEDs). A 20-month on-the-ground study in 2014-2015 by the European Parliament’s Conflict Armament Research (CAR) implicated 50 commercial entities that supplied materials used for homemade explosives to IS from 20 different countries. “With 13 companies involved in

Islamic State Networks in Turkey

Page 15

the supply chain,” the report read, “Turkey is the most important choke point for components used in the manufacture of IEDs by IS forces.”71

In addition to explosive agents, reports of foreign fighters crossing in and out of Turkey have been commonplace. Indeed, according to more than 4,600 IS documents reviewed by the U.S. army, more than 90 percent of all foreign fighters – including Turkish militants – between 2013 and 2014 entered Syria through the Turkish border. Most fighters flew to Istanbul, traveled from there to the border provinces of Sanliurfa and Gaziantep, and then to transit points in Kilis, Elbeyli, Karkamis, or Akcakale.72 In nearly every media interview of former jihadists, the military and border guards are described as unconcerned about the flow of fighters. In some, they are described as openly complicit.73

Turkish IS Militants in Syria and Iraq

The majority of foreign fighters that have joined IS are believed to hail primarily from Europe or Arab states.74 However, since 2012, a steady stream of jihadist Turks have taken up arms to join the war in Syria.75 In early 2015, then-Deputy Prime Minister Bulent Arinc declared that around 1,000 Turkish citizens had

71. “Tracing the Supply of Components Used in Islamic State IEDs,” Conflict for Armament Research, February 2016, page 11. (http://www.conflictarm.com/publications/)72. Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)73. Katrin Elger, Hasnain Kazim, Christoph Reuter, and Holger Stark, “The Caliphate Next Door: Turkey Faces Up to its Islamic State Problem,” Spiegel (Germany), September 29, 2014. (http://www.spiegel.de/international/world/islamic-state-jihadist-activity-in-northern-syria-forces-Turkey-hand-a-994392.html)74. “The Foreign Fighters Phenomenon in the European Union,” International Centre for Counter-Terrorism – The Hague, April 2016. (https://www.icct.nl/wp-content/uploads/2016/03/ICCT-Report_Foreign-Fighters-Phenomenon-in-the-EU_1-April-2016_including-AnnexesLinks.pdf )75. “Halep’te 50 Türk savaşıyor (50 Turks Are Fighting in Aleppo),” Radikal (Turkey), August 25, 2012. (http://www.radikal.com.tr/dunya/halepte-50-turk-savasiyor-1098170/)76. “Bülent Arınç: Türkiye’den IŞİD’e katılan gençlerin sayısı 1000 civarında (Bülent Arınç: Number of Turkish youth joining IS is around 1000),” Radikal (Turkey), February 23, 2015. (http://www.radikal.com.tr/politika/bulent-arinc-turkiyeden-iside-katilan-genclerin-sayisi-1000-civarinda-1299255/)77. Serhat Erkman, “Suriye ve Irak’ta Savaşan Türkiyeli Mücahitler (Turkish Foreign Fighters in Syria and Iraq),” Al Jazeera Turk (Turkey), May 25, 2015. (http://www.aljazeera.com.tr/gorus/suriye-ve-irakta-savasan-turkiyeli-mucahitler)78. Ibid.79. Ipek Yezdani, “Türkiye’den cihada 10 bin kişi gitti (10,000 people from Turkey have gone to jihad),” Hurriyet Daily News (Turkey), June 27, 2015. (http://www.hürriyet.com.tr/turkiye-den-cihada-10-bin-kisi-gitti-29383934)

joined IS.76 Hurriyet reported a higher number that year, noting that 2,700 Turks had joined the group and around 1,500 had returned to Turkey.77 According to one account, however, the number of Turks who traveled to Syria and Iraq over the last four years – including militants, migrant families, those killed on the battlefield, and those who have returned to Turkey – could be up to 10,000.78

Admittedly, some families have simply “migrated” to IS-controlled territory with the sole purpose of residing under the caliphate. According to Turkish reports, families make up around 60 percent of those travelers to Syria and Iraq included in the 10,000-estimate.79 Many others, meanwhile, have directly gone to fight in battles on behalf of IS, even assuming leadership positions with the group.

The Kurdish Factor

Turkish nationals, both Turks and Kurds, have joined groups like the Islamic State and Nusra Front for another reason: to battle the PKK and its affiliates, which are all designated terrorist groups in Turkey. The PKK’s Syrian wing, the Democratic Union Party (PYD), and its armed wing, the Peoples’ Protection Units (YPG), have vied for self-governance in northern

Islamic State Networks in Turkey

Page 16

Syria since 2012. Turks have joined jihadist groups to fight the PYD/YPG out of a sense of religious or national duty, sometimes both.

The PYD has blamed Turkey for either assisting these jihadi groups in their attacks against the Kurds or turning a blind eye to them. And it is not hard to understand why Turkey might have adopted this strategy. Ankara has long considered Kurdish separatism the top threat to its territorial sovereignty. Both the PYD and PKK claim to represent millions of Kurds residing in southern Turkey and northern Syria, divided by the border but connected by lineage. An autonomous Kurdish region and a strengthened PYD on the Syrian side of the border could inflame Kurdish nationalism on the Turkish side, not to mention a resurgent PKK.

Tensions soared in the fall of 2014, when the Islamic State’s jihadists sieged the Syrian-Kurdish town of Kobani, and the Turkish military merely watched from across the border.80 Finally, after the U.S. air dropped ammunition for the Kurds under siege, Turkey allowed the passage of Iraqi Kurdish peshmerga fighters – who are unaffiliated with the PKK – into Kobani to help their Syrian brethren there.81

The YPG’s liberation of Kobani in January 2015 was a turning point for both Turkey and the U.S. Washington’s decision to partner with the YPG in the fight against IS in Syria set off alarm bells in Ankara. That summer, Erdogan, to the shock of many, declared the Dolmabahce agreement, the cornerstone for Turkey’s Kurdish peace process, null and void.82 It did not

80. Dexter Filkins, “When Bombs Aren’t Enough,” The New Yorker, October 10, 2014. (http://www.newyorker.com/news/daily-comment/turkey-kurds-battle-isis-kobani); Henri J. Barkey and Eric S. Edelman, “The Obama Administration, Turkey and Syria’s Kurds,” The American Interest, October 7, 2014. 81. Dasha Afanasieva, “Iraqi peshmerga fighters arrive in Turkey for Syria deployment,” Reuters, October 29, 2014. (http://uk.reuters.com/article/uk-mideast-crisis-peshmerga-idUKKBN0II0P720141029) 82. “Erdoğan’s denial of ‘Dolmabahçe Agreement’ sparks row,” Hurriyet Daily News (Turkey), July 20, 2015. (http://www.hürriyetdailynews.com/erdoğans-denial-of-dolmabahce-agreement-sparks-row.aspx?pageID=238&nID=85656&NewsCatID=338)83. Rustem Erkan, “Güneydoğu’da seçim dinamikleri (Voting Dynamics in the South East),” Radikal (Turkey), December 14, 2008. (http://www.radikal.com.tr/radikal2/guneydoguda_secim_dinamikleri-912863); “1990’larda Kürt siyasi hareketi (Kurdish Politics in the 1990s),” Al Jazeera Turk (Turkey), December 26, 2013. (http://www.aljazeera.com.tr/dosya/1990larda-kurt-siyasi-hareketi)84. Mahmut Bozarslan, “Kurds fight Kurds in Syria,” Al Monitor, June 25, 2015. (http://www.al-monitor.com/pulse/en/originals/2015/06/turkey-syria-kurdish-is-militants-fight-brethren.html)

take long before the decades-long war resumed. Since July 2015, Turkey has carried out numerous airstrikes against the PKK in northern Iraq, and the YPG has emerged as its top enemy in Syria. In August 2016, Turkish troops entered the IS-held town of Jarabulus to clear IS from its border, as well as to establish a protected zone between two Kurdish enclaves to prevent the Kurds from linking their territories into a contiguous entity.

The IS-Kurdish conflict in Iraq and Syria, meanwhile, has itself been spilling over to Turkey as well. Some of the most deadly Islamist factions in Turkey, such as Kurdish Hizbullah (Huda-Par), have emerged from ultraconservative provinces in the country’s predominantly Kurdish southeast.83 Clashes between rival PKK- and Huda-Par-affiliated Kurds in Turkey during the siege of Kobani in 2014 was one of the early signs of this spillover.

It is unclear whether Huda-Par, which has a support base of about 100,000, actually supports the terror group. But secular Kurds have reason to worry. At least 600 of Turkey’s Kurds have reportedly joined IS in Syria and Iraq since 2014.84

Turkey and IS at War

The Turkish-IS relationship has grown increasingly hostile since Turkey officially declared war on the group in mid-2015. In addition to Ankara permitting the U.S.-led anti-IS coalition to use Turkey’s Incirlik airbase, Turkish artillery and warplanes began bombarding the group’s positions along the border.

Islamic State Networks in Turkey

Page 17

Six of the ten major IS attacks in Turkey have taken place since that summer, four of which are among the deadliest terror attacks in Turkish history and the other two directly targeting tourists in Istanbul – mostly German and Israeli.85 Between January and May 2016, IS fired more than 70 rockets into Turkey’s border town Kilis (rockets fueled by the aforementioned Turkish-made potassium nitrates).86

The hostility took an even sharper turn in August 2016, when Turkish troops entered Syria for “Operation Euphrates Shield,” a move primarily to counter the Kurdish territorial gains in northern Syria but also aimed at targeting IS positions along the way. In October, the Turks and their Arab-Syrian anti-regime allies began to advance towards the IS-held city of al-Bab in northwest Syria. In response, the Islamic State’s “Caliph” Abu Bakr al-Baghdadi personally called for attacks against Turkey for the first time in early November.87 A month later, a grotesque video emerged showing two soldiers, allegedly Turkish, burned alive by the group.88 In a trademark Turkish move, Ankara tried to bury the scandal by blocking access to YouTube, Twitter, and Facebook.89 Harder to hide, however, are the dozens of Turkish soldiers so far killed by IS inside Syria since the advance to al-Bab.90

85. See Appendix I. 86. Selin Girit, “Syria conflict: Kilis, the Turkish town enduring IS bombardment,” BCC (UK), May 9, 2016. (http://www.bbc.com/news/world-europe-36245505)87. “Daesh terrorist leader Baghdadi calls for attacks against Turkey,” Daily Sabah (Turkey), November 3, 2016. (http://www.dailysabah.com/mideast/2016/11/03/daesh-terrorist-leader-baghdadi-calls-for-attacks-against-turkey)88. “IS’ Aleppo Province Releases Video of Burning to Death Two Soldiers in Turkish Army,” SITE Intelligence Group, December 22, 2016. (https://news.siteintelgroup.com/Jihadist-News/is-aleppo-province-releases-video-of-burning-to-death-two-soldiers-in-turkish-army.html)89. Caroline O’Donnovan, “Social Media Reportedly Blocked In Turkey After Horrific ISIS Video,” BuzzFeed, December 22, 2016. (https://www.buzzfeed.com/carolineodonovan/social-media-reportedly-blocked-in-turkey-after-horrific-isi?utm_term=.nomoppO4m#.mcv3aa1kg)90. Sarah El Deeb, “Turkey bogged down in Syria as it realigns with Russia,” Associated Press, January 11, 2017. (https://apnews.com/4b50b17b41fb4ba5a6c70fd1d746f753/Turkey-bogged-down-in-Syria-as-it-realigns-with-Russia)91. Seval Guler, “Turkey foiled 339 terror plots in 2016: Interior min.,” Anadolu Agency (Turkey), January 3, 2017. (http://aa.com.tr/en/todays-headlines/turkey-foiled-339-terror-plots-in-2016-interior-min/719684)92. “Turkey: 820 people detained in operations against Daesh,” Anadolu Agency (Turkey), February 6, 2017. (http://aa.com.tr/en/todays-headlines/turkey-820-people-detained-in-operations-against-daesh/742315) 93. “ISIS Renting Villas in Istanbul,” Aydinlik (Turkey), September 29, 2014. (http://aydinlikdaily.com/Detail/ISIS-Renting-Villas-In-Istanbul/4619#.VClv7xuHGzE.twitter) 94. “Looking for ISIL: How jihadists operate among Turks,” Hurriyet Daily News (Turkey), September 22, 2014. (http://www.Hürriyetdailynews.com/looking-for-isil-how-Turks-become-jihadists.aspx?pageID=238&nID=72054&NewsCatID=352)

Islamic State Networks in Turkey

With IS attacks against Turkey clearly surging, Ankara significantly widened its intelligence-gathering efforts and crackdowns on local Islamic State suspects. The Interior Ministry announced that Turkish security forces prevented 22 IS-linked “major terror incidents” in 2016.91 It also said that 1,338 Islamic State suspects, of whom 694 were foreign nationals, were detained throughout the year. In early February of this year, authorities detained 820 IS suspects in a massive two-day operation that spanned 29 Turkish provinces.92

Ankara’s recent efforts are commendable, but they are too little and too late in coming. After years of permissive border policies, Turkey is deeply exposed to the jihadist threat. The AKP government’s initial indifference to Islamic State activities allowed the group to establish its networks inside Turkey, raise funds, and recruit Turkish jihadists.93 Throughout the fall of 2014, Turkish newspapers reported often on IS activities in large cities across IS-held towns, such as Gaziantep and Sanliurfa, as well as in more remote provinces such as Konya in central Anatolia.94

Islamic State Networks in Turkey

Page 18

Reports indicate that Islamic State recruitment in Turkey continued despite the concern.95

According to Turkish police and security units’ reports from July 2016, a 60-person core cadre inside Turkey is operating along the Istanbul-Sanliurfa-Gaziantep-Hatay-Batman-Adiyaman-Kahramanmaras area. This core group has been professionally trained abroad for establishing cells and planning attacks.96

There are undoubtedly more operatives in the country, and the pool for recruits are at worrying levels. According to an October 2014 poll by the Turkish research center Metropoll, nearly two percent of respondents said they knew people who joined the Islamic State. Turkey’s population is nearly 80 million. That means 1.6 million Turks today may personally know IS members. Even if only one percent of those are sympathetic to the jihadist cause, that would mean 16,000 individuals have the will and personal connections to join or recruit for the Islamic State.

The geographic distribution of detained and arrested IS suspects, meanwhile, give a glimpse of how widespread the group’s network is and how mobile its militants are. Of the four Turkish provinces bordering Syria, three – Gaziantep, Sanliurfa, and Kilis – shared a border with

95. Alev Scott and Alexander Christie-Miller, “Exclusive: ISIS Starts Recruiting in Istanbul’s Vulnerable Suburbs,” Newsweek, September 12, 2014. (http://www.newsweek.com/2014/09/19/exclusive-how-istanbul-became-recruiting-ground-islamic-state-269247.html); Ceylan Yeginsu, “ISIS Draws a Steady Stream of Recruits From Turkey,” The New York Times, September 15, 2014. (http://www.nytimes.com/2014/09/16/world/europe/turkey-is-a-steady-source-of-isis-recruits.html?_r=0)96. Metehan Demir, “Rakamlarla IŞİD’in Türkiye’de tüyler ürperten varlığı (The chilling presence of ISIS in Turkey, by the numbers),” Super Haber (Turkey), July 1, 2016. (http://www.superhaber.tv/dunya/rakamlarla-isidin-turkiyede-tuyler-urperten-varligi/haber-10824)97. See Appendix II.98. Alican Uludag, “70 ilde IŞİD hücresi var: CHP ve HDP hedefte (ISIL cells in 70 provinces: CHP and HDP are targets),” Cumhuriyet (Turkey), April 12, 2016. (http://www.cumhuriyet.com.tr/haber/turkiye/514593/70_ilde_ISiD_hucresi_var__CHP_ve_HDP_hedefte.html)99. Noah Blaser, Aaron Stein, and Has Avrat, “Islamic State in Turkey: Network Charts and Detention Maps,” North Caucasus Caucus, April 6, 2016. (http://northcaucasuscaucus.blogspot.com.tr/?m=1); Humeyra Pamuk and Orhan Coskun, “Investigators focus on home-grown Islamic State cell in Turkey bomb probe,” Reuters, October 13, 2015. (http://www.reuters.com/article/us-turkey-explosion-investigation-idUSKCN0S72AO20151013)100. Piotr Zalewski, “The Kurdish Town That Hosts Turkey’s Deadliest Terror Cell,” Financial Times (UK), October 18, 2015. (http://www.ft.com/intl/cms/s/0/689f9fbe-741b-11e5-bdb1-e6e4767162cc.html#axzz4Bm9bSSpw); Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)101. Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)

the Islamic State between 2014 and 2015. The fourth, Hatay, sits across from Idlib, also a jihadist stronghold in northwest Syria. In addition to these four border provinces, Turkish authorities have carried out anti-IS operations in at least 50 other provinces across the country in 2016.97 According to a classified 2015 Turkish police report leaked last year, there could be IS cells in up to 70 of Turkey’s 81 provinces.98

Interrogations of detained Islamic State members point to multiple local Turkish IS units established throughout Turkey.99 The deadliest, known as “the Adiyaman cell,” established in 2013 and led by Mustafa Dokumaci, has produced multiple perpetrators of IS suicide attacks in Turkey.100 Three IS-linked suicide bombings (two of them twin-bombings) since June 2015 were committed by Turkish citizens, and all of the four known perpetrators – Orhan Gonder, Seyh Abdurraman Alagoz, Yunus Emre Alagoz, and Omer Deniz Dundar – were from this unit.101 Along with Mustafa Dokumaci, the group’s top recruiter is reportedly Ahmet Korkmaz, and other key operational

“ According to a classified 2015 Turkish police report leaked last year, there could be IS cells in up to 70 of Turkey’s 81 provinces.”

Islamic State Networks in Turkey

Page 19

members are Ibrahim Bali, Yunus Durmaz, and Halil Ibrahim Durgun.102 While Bali is said to have been responsible for the cell’s Turkey-based operations, he reportedly moved to Raqqa around July and was replaced by Mustafa Mol from Sanliurfa.103

Along with the Adiyaman network, the Bingol cell is another homegrown IS unit located in a conservative Kurdish-majority town with a strong history of Islamism.104 According to Turkish media reports, other major recruitment centers appear to have been established in Istanbul, Izmir, Konya, as well as Gaziantep.105 The network in Gaziantep has seen numerous anti-IS police raids in the last year, with large quantities of suicide vests confiscated in the operations.106

The Adiyaman cell is testament to the security and intelligence failures of the Turkish government. The two Alagoz brothers owned a teahouse in the southern

102. “Turkey’s Jihadists,” Turkey Wonk: Nuclear and Political Musings in Turkey and Beyond, April 28, 2015. (https://turkeywonk.wordpress.com/2015/04/28/turkeys-jihadists/)103. Tolga Sardan, “IŞİD’in yeni celladı: MOL (ISIL’s New Executioner: MOL),” Milliyet (Turkey), July 9, 2016. (http://www.milliyet.com.tr/isid-in-yeni-celladi-mol-gundem-2274994/)104. Mahmut Bozarslan, “Kurds fight Kurds in Syria,” Al Monitor, June 25, 2015. (http://www.al-monitor.com/pulse/en/originals/2015/06/turkey-syria-kurdish-is-militants-fight-brethren.html)105. For more on the recruitment activities of the Adıyaman, Istanbul, Konya, and Izmir cells, see: Aaron Stein, “Islamic State Networks in Turkey: Recruitment for the Caliphate,” The Atlantic Council, October 2016. (http://www.publications.atlanticcouncil.org/islamic-state-networks-in-turkey/)106. “ISIL member blows himself up during Turkish police raid,” Al Jazeera, May 20, 2016. (http://www.aljazeera.com/news/2016/05/isil-member-blows-turkish-police-raid-160520090343990.html); Fevzi Kizilkoyun, “ISIL militants prepared suicide vests in Gaziantep on Syrian border: Police,” Hurriyet Daily News (Turkey), December 5, 2015. (http://www.hürriyetdailynews.com/isil-militants-prepared-suicide-vests-in-gaziantep-on-syrian-border-police.aspx?pageID=238&nID=92101&NewsCatID=509) 107. “İşte IŞİD bombacılarının yollarının kesiştiği Adıyaman›daki İslam Çay Ocağı (Here is the tea house in Adıyaman where ISIL bombers crossed paths),” CNN Türk (Turkey), July 23, 2015. (http://www.cnnturk.com/turkiye/iste-isid-bombacilarinin-yollarinin-kesistigi-Adıyamandaki-islam-cay-ocagi)108. “Suruç bombacısının ağabeyinin açtığı çay ocağı Adıyaman’daki IŞİD bürosu muydu? (Was the tea house that the Suruç bomber’s brother opened ISIL’s bureau in Adıyaman?),” Radikal (Turkey), July 23, 2015. (http://www.radikal.com.tr/turkiye/Suruç-bombacisinin-agabeyinin-actigi-cay-ocagi-Adıyamandaki-isid-burosu-muydu-1401949/)109. Aki Peritz, “The Islamic State Has Sleeper Cells Throughout Turkey. Does Erdoğan Care?” The New Republic, January 21, 2015. (https://newrepublic.com/article/120821/islamic-state-isis-has-sleeper-cells-turkey-says-national-police)110. Noah Blaser, Aaron Stein, and Has Avrat, “Islamic State in Turkey: Network Charts and Detention Maps,” North Caucasus Caucus, April 6, 2016. (http://northcaucasuscaucus.blogspot.com.tr/?m=1)111. “Turkey vs. ISIS and the PKK: What’s the Difference?” Bipartisan Policy Center, December 9, 2016. (http://bipartisanpolicy.org/turkey-isis-pkk-maps/#1467822414183-2ca6b3ce-572c)112. Metin Gurcan, “Are Turkey’s efforts to combat foreign fighters too late?” Al Monitor, July 12, 2016. (http://www.al-monitor.com/pulse/originals/2016/07/turkey-too-late-dealing-foreign-terrorist-fighters.html)

Turkish city of Adiyaman, where they recruited young men.107 The families of these new targets repeatedly notified the police throughout 2014, urging for action against the Islamic State recruiters. The authorities did not shut the teahouse down until the end of the year.108

According to a Turkish intelligence report disclosed by Jane’s Intelligence Weekly in January 2015, up to 3,000 IS sleeper cells were suspected to be in Turkey.109 But the number of IS-related arrests at that time was appallingly low. According to one tally, only 127 Islamic State suspects were detained between 2014 and the summer of 2015.110 That number jumped significantly after the July 2015 suicide attack in Suruc, reaching nearly 400 by the following July,111 and more than 1,300 by the end of the year.112

The increasing number of detentions, however, is misleading: Many IS suspects were released within

Islamic State Networks in Turkey

Page 20

days or weeks of detention.113 Following a massive anti-IS operation in Konya in December, for example, all 43 detainees were released within a week.114 While Turkey boasted that it detained 1,338 IS-related suspects in 2016, the number of convictions stood at a mere seven.115

One problem for Ankara is that the legal standards for convicting jihadists in Turkey are too high. Turkish foreign fighters who return from IS battlegrounds in Iraq or Syria are not convicted unless they are proven to have directly attacked the Turkish homeland or its citizens. One former Islamic State executioner, according to his July 2015 interview with the Turkish daily BirGun, is now living freely in Ankara, working as a parking lot attendant. Another IS assassin, caught on camera shooting a man in Syria’s Latakia, received a reduced jail sentence due to “good conduct” last July.116

Islamic State members who have carried out attacks against Turkey, meanwhile, are not sentenced immediately – nor to the fullest extent of the law.117 The perpetrators of the March 2014 Islamic State attack in

113. “IŞİD Davasının ilk Duruşmasında 12 Sanık Tahliye Edildi (12 Suspects Released in the First Hearing of ISIL Case),” Hurriyet (Turkey), October 25, 2016. (http://www.hurriyet.com.tr/isid-davasinin-ilk-durusmasinda-12-sanik-tahliy-40258833)114. “IŞİD Operasyonunda Gözaltına Alınan 43 Kişi Serbest Bırakıldı (43 People Detained in ISIL Operation Released),” Sol (Turkey), December 23, 2016. (http://haber.sol.org.tr/toplum/isid-operasyonunda-gozaltina-alinan-43-kisi-serbest-birakildi-180036)115. “Bakanlık: Cezaevlerinde 513 IŞİD üyesi Var (Ministry: There are 513 ISIL Members in the Prisons),” Diken (Turkey), December 28, 2016. (http://www.diken.com.tr/adalet-bakanligi-cezaevlerinde-513-isid-uyesi-var/)116. “ISIL militant’s jail sentence reduced due to ‘good conduct’ in Turkey,” Hurriyet Daily News (Turkey), July 11, 2016. (http://www.hürriyetdailynews.com/Default.aspx?pageID=238&nID=101468&NewsCatID=509)117. Metin Gurcan, “Are Turkish courts delaying trials of IS militants?” Al Monitor, July 31, 2015. (http://www.al-monitor.com/pulse/en/originals/2015/07/turkey-isis-detentions-questions-remains-legal-system.html)118. “Niğde saldırısı faili IŞİD’lilere toplam 10 kez ağırlaştırılmış müebbet (10 times aggravated sentence for ISIL-member Nigde attack perpetrators),” Diken (Turkey), June 15, 2016. (http://www.diken.com.tr/nigde-saldirisi-faili-isidlilere-toplam-10-kez-agirlastirilmis-muebbet/)119. “IŞİD’in karar çıkması beklenen Niğde saldırısı davası 17 Mart’a ertelendi (ISIL’s Nigde attack trial expected to issue conviction delayed to March 17),” CNN Türk (Turkey), January 20, 2016. (http://www.cnnturk.com/turkiye/isidin-karar-cikmasi-beklenen-nigde-saldirisi-davasi-17-marta-ertelendi)120. “ISIL militant linked to Istanbul nightclub attacker says he was previously detained, released,” Hurriyet Daily News (Turkey), February 2, 2017. (http://www.hürriyetdailynews.com/isil-militant-linked-to-istanbul-nightclub-attacker-says-he-waspreviouslydetained-released.aspx?pageID=238&nID=109272&NewsCatID=509)121. “Askerlere infaz iddiasını içeren Videodaki IŞİD’li iki [Defa] Gözaltına Alınıp Serbest Bırakılmış (The ISIL Member in the Video Depicting Execution of Soldiers Was Detained and Released Twice),” T24 (Turkey), January 21, 2017. (https://t24.com.tr/haber/askerlere-infaz-iddiasini-iceren-videodaki-isidli-iki-gozaltina-alinip-serbest-birakilmis,384385)122. “Seven ISIL Suspects Released By Istanbul Court,” Hurriyet Daily News (Turkey), March 24, 2016. (http://www.hürriyetdailynews.com/seven-isil-suspects-released-by-istanbul-court.aspx?PageID=238&NID=96893&NewsCatID=509)

central Anatolia, for example, were not convicted until 2016.118 The trial was postponed several times, and the judge was replaced four times.119

Additionally, the same suspects often get detained and released multiple times. That list includes Ilyas Mamasharipov, a Tajik Islamic State member who ordered the Reina nightclub attack in Istanbul. Upon his arrest on January 31, he testified that he had been detained and sent to a deportation center in Turkey, but was ultimately released.120 There is also the case of Hasan Aydin, one of the Turkish IS militants from the December video depicting the burning of two Turkish soldiers in Syria. Aydin was reportedly detained in Adana in 2012, and then again in Hatay in 2015 for trying to cross into Syria.121

There is also the case of Halis Bayancuk (aka Abu Hanzala), Turkey’s most notorious IS advocate,122 who

“ While Turkey boasted that it detained 1,338 IS-related suspects in 2016, the number of convictions stood at a mere seven.”

Islamic State Networks in Turkey

Page 21

was arrested once in 2008 for planning a synagogue attack in Istanbul, and twice later for his affiliation with IS – first in January 2014 and then again in July 2015. He was part of a 96-person Islamic State suspect list in a trial which, after its fourth hearing in March 2016, released all 96 suspects.123

This lax enforcement stands in stark contrast to the treatment of Kurdish nationalists and other opposition figures in Turkey. Kurdish politicians and civilians are often detained and convicted under dubious terrorism charges.

Online Radicalization: Jihadists on Turkish

Social MediaIn 2007, Turkey’s Savings Deposit Insurance Fund took over 63 media organizations, including the country’s second-largest media conglomerate Sabah-ATV.124 This was the AKP’s first blatant attempt at controlling the media: Columnists critical of the ruling party were promptly dismissed and replaced with loyalists. By the following spring, Sabah-ATV was sold to a holding company owned by then-Prime Minister Erdogan’s close associate and run by the premier’s son-in-law.125 A year later, Erdogan went after Turkey’s largest media

123. “IŞİD Davası’nda tüm sanıklar tahliye edildi (All suspects in ISIL case released),” Cumhuriyet, March 24, 2016. (http://www.cumhuriyet.com.tr/haber/turkiye/503898/ISiD_Davasi_nda_tum_saniklar_tahliye_edildi.html)124. “TMSF atv ve Sabah’a el koydu (TMSF seized ATV and Sabah),” Hurriyet Daily News (Turkey), April 2, 2007. (http://www.hürriyet.com.tr/tmsf-atv-ve-sabaha-el-koydu-6248828) 125. “Circulation wars,” The Economist, May 8, 2008. (http://www.economist.com/node/11332305)126. Asli Aydintasbas, “Turkey’s War on the Press,” The Wall Street Journal, September 18, 2009. (http://www.wsj.com/articles/SB10001424052970203917304574414503346981992)127. “Turkey is Facebook world country No. 4,” Social Bakers, August 23, 2011. (https://www.socialbakers.com/blog/207-turkey-is-facebook-world-country-no-4)128. Neal Caren and Didem Türkoğlu, “Political controversy has helped fuel Twitter’s growth in Turkey,” The Washington Post, March 28, 2014. (https://www.washingtonpost.com/news/monkey-cage/wp/2014/03/28/political-controversy-has-helped-fuel-twitters-growth-in-turkey/)129. Ibid.130. “Raid on ‘Occupy Taksim Park’ demonstrators triggers outcry,” Hurriyet Daily News (Turkey), May 30, 2013. (http://www.hürriyetdailynews.com/raid-on-occupy-taksim-park-demonstrators-triggers-outcry.aspx?pageID=238&nID=47870&NewsCatID=341)131. Pelin Turgut, “As Turkey’s Protests Continue, Attention Falls on Failures of Turkish Media,” TIME, June 6, 2013. (http://world.time.com/2013/06/06/as-turkeys-protests-continue-attention-falls-on-failures-of-turkish-media/)132. Olga Khazan, “These Charts Show How Crucial Twitter Is for the Turkey Protesters,” The Atlantic, June 12, 2013. (http://www.theatlantic.com/international/archive/2013/06/these-charts-show-how-crucial-twitter-is-for-the-turkey-protesters/276798/)

conglomerate, the Dogan Media Group, slapping it with a staggering $2.5-billion tax fine.126

Ankara’s steady dismantling of traditional media outlets, as well as its systematic intimidation of others, fueled the rise of a highly polarized Turkish social media. By 2011, Turkey had become Facebook’s fourth largest market worldwide with almost 30 million Turkish accounts – behind only the United States, India, and Indonesia.127 Twitter launched its Turkish-language version that same year, and the website’s usage reportedly “exploded.” According to The Washington Post, nearly 200,000 Turkish internet users signed up each month in 2011.128 By the next year, that figure had nearly doubled.129

In May 2013, the Turkish police’s decision to disperse a small environmentalist sit-in at Istanbul’s Gezi Park with brute force unleashed a massive, nationwide anti-government movement that lasted months.130 As deadly clashes rocked Taksim Square on June 2, Turks tuned to their country’s largest independent television channel, CNN Turk, only to find a 3-hour documentary on penguins.131 With the exception of one channel, Halk TV, Turks had only Twitter, Facebook, and blogs to keep up with the events. Twitter thus emerged as the most important tool for protesters to communicate and mobilize.132 It was during these protests that

Islamic State Networks in Turkey

Page 22

Erdogan infamously called Twitter “the worst menace to society.”133

Today, most traditional media outlets in Turkey are either directly controlled by or under the pressure of the AKP.134 Of course, the AKP has had the power to unilaterally block websites for the last decade, but since April 2015, it has also been legally allowed to block them for political reasons, such as defamation. Ankara has now blocked over 100,000 websites.135 In November 2016, Freedom House downgraded Turkey’s internet freedom rating to “not free,” citing periodic social media bans and the internet blackout in some Kurdish areas in Turkey’s east.136 As the Reuters Institute for the Study of Journalism notes, this suppression of print media has turned blogs and social media into the most important news platforms in Turkey.137

The Islamic State’s Turkish-Language Activity on Twitter

With social media usage skyrocketing in Turkey, the Islamic State has taken note. The caliphate has consistently used Turkish-language messages in its outreach campaign. In its online Turkish-language magazine, IS has threatened Turkey with conquest and vilified its leaders as secular and un-Islamic. The magazine is called Konstantiniyye, the Ottoman word for Constantinople, now Istanbul. Its first issue featured a modern cover photo of the city,

133. “Erdoğan: ‘social media is the worst menace to society,’” France 24 (France), June 4, 2013. (http://www.france24.com/en/20130603-turkey-twitter-social-media-menace-erdoğan-sarkozy-london-goldman-sachs-parallel-diplomacy) 134. Claire Sadar, “Censorship Scandal Leads to Resignation of Head of Turkish Media Conglomerate,” Muftah.org, October 2, 2016. (http://muftah.org/censorship-scandal-leads-resignation-head-turkish-media-conglomerate/#.WFr-KFMrKM) 135. The figure is from EngelliWeb.com, a project created in 2008 to track internet censorship in Turkey, accessed in November 2016. As of February 2017, the website is no longer active. 136. “Turkey rated ‘not free’ in internet freedom in new Freedom House report,” Hurriyet Daily News (Turkey), November 15, 2016. (http://www.hürriyetdailynews.com/turkey-rated-not-free-in-internet-freedom-in-new-freedom-house-report.aspx?pageID=238&nid=106140)137. Nic Newman, Richard Fletcher, David A. L. Levy, and Rasmus Kleis Nielsen, “Digital News Report 2016,” Reuters Institute for the Study of Journalism, accessed March 21, 2017. (http://www.digitalnewsreport.org/)138. Bradley Klapper, “AP: Islamic State’s Twitter traffic drops amid US efforts,” Associated Press, July 9, 2016. (http://bigstory.ap.org/article/21c9eb68e6294bdfa0a099a0632b8056/ap-exclusive-islamic-states-twitter-traffic-plunges)139. Yasmeen Serhan, “Twitter’s Crackdown on Users Promoting Terrorism,” The Atlantic, August 18, 2016. (http://www.theatlantic.com/news/archive/2016/08/twitter-suspends-users-terrorism/496481/?utm_source=atltw)140. Suleyman Ozeren, Hakan Hekim, Salih Elmas, and Halil Ibrahim Canbegi, “ISIS in Cyberspace: Findings From Social Media Research,” Global Policy and Strategy Institute, February 2016. (https://www.academia.edu/24311407/ISIS_in_Cyberspace_Findings_From_Social_Media_Research)

and implied its intention to conquer it. The magazine’s second issue made it personal, denouncing Turkish President Erdogan as tagut – a tyrant.

Since 2015, IS’s overall Twitter traffic has reportedly dropped by 45 percent, thanks to U.S. efforts for countering jihadi propaganda online.138 Twitter has also increased its scrutiny over jihadist propaganda, suspending more than 360,000 accounts between 2015 and August 2016.139 These are improvements, but both the U.S. government and Twitter’s efforts have, naturally, focused mostly on the group’s Arabic- and English-language accounts. The jihadists and their sympathizers continue to promote their messages in Turkish with little intervention.

One study by the Ankara-based Global Policy and Strategy Institute analyzed 25,403 tweets posted by 290 pro-IS Turkish-language accounts over the course of July 2015 – the same month as the Suruc attack in southern Turkey.140 One month’s worth of tweets repeated words such as jihad, idolatry, excommunication, kill, assault, and weapon more than 30,000 times. Their targets were not only westerners, Christians, and Jews, but also “infidels” who defy IS’s version of Islam, such as Kurds and secular Turks. Tweets conveyed news directly from the battlefield, recruitment announcements, or smuggling opportunities. Others sought to publicize IS’s worldview and ideology.

Islamic State Networks in Turkey

Page 23

A similar study from June 2015 by Ankara’s TOBB University captured 51,000 Twitter messages using 52 keywords, and identified 21 key accounts with a network of 2,567 individuals engaged in pro-IS messaging, of which 87.7 percent were tweeting from outside of Turkey.141 Monitoring those accounts for over a year, researchers studied IS’s overt propaganda and recruitment efforts. Notably, the study found the top agenda topic of pro-IS accounts to be the PKK and the Kurds. The IS-PYD clashes in Syria, the study found, translated into intensified IS recruitment on Turkish Twitter.

Sidar Global Advisors' Study of IS on Turkish-Language Twitter

In 2016, FDD commissioned its own study into the Islamic State and jihadist accounts on Turkish Twitter. The study was conducted by Sidar Global Advisors (SGA), which used Brandwatch Analytics, a social media monitoring and analytics platform, to collect Turkish tweets over a 45-day period between March 15 – four days before the IS attack in Istanbul’s Sultanahmet district – until May 1.142

The study identified 137 key IS-sympathizer accounts and, searching 18 keywords, analyzed 693,533 mentions on Twitter that were analyzed for topic, sentiment, author, geo-location, and gender. In line with the findings of other similar studies, few tweets were found to be geo-tagged. A geo-locational analysis based on those tweets, however, showed most of them generating from Turkey’s three major cities – Istanbul, Ankara, and Izmir – with some coming from the country’s east. Outside of Turkey, Turkish-language tweets were recorded in Belgium, the Netherlands, and Germany.

The three most-used keywords were cihad (jihad), kafir (infidel), and Nusra (a reference to al-Qaeda’s Syrian branch). In addition, Brandwatch often picked up mentions of Kudus (Jerusalem), mucahid (mujahid), and siyonist (Zionist). Indeed, proponents of jihadism

141. “İşte Twitter›daki Türk IŞİD›çilerle ilgili çarpıcı detaylar (Striking details about Turkish pro-ISIS accounts on Twitter),” Hurriyet Daily News (Turkey), September 22, 2016. (http://www.hürriyet.com.tr/iste-twitterdaki-turk-isid-sempatizanlarinin-gundemi-40229219)142. For the full report of the SGA study, see Appendix III.

on Turkish Twitter appeared to be animated by the Palestinian-Israeli conflict, while adopting other common words and themes in the jihadi lexicon.

Many of the accounts we recorded were shut down by Twitter throughout the period of the study, with the authors shortly reemerging on the platform under similar account names. The account of a suspected IS militant by the Twitter name of Ali Demircioglu, for example, was suspended and recreated several times under similar names. The user’s last known account was @1ali2demir2, whose profile picture was of a black-clad figure drawing his sword. This account has also been suspended since.

SGA’s study recorded spikes in jihadi language on Turkish Twitter during or after three types of events: IS-linked terror attacks, important foreign dignitary visits, and operational developments in Syria.

Among the recorded Turkish-language tweets with jihadist sympathies, two themes stood out in particular: the PKK and the Assad regime, whose respective atrocities against the Turks and Syrians were compared to the Islamic State’s attacks against Europeans. Following the Islamic State bombings in Brussels, for example, @1ali2demir2 shared a picture from the attack scene with the Turkish caption: “You should have protected your airport instead of the atheist PKK terror organization’s tent” – implying that the Europeans were backing the Kurdish terrorist group.

One of the most influential accounts that the study tracked, @Latormenta, also now defunct, tweeted about the Brussels attacks on March 22 with nearly 100,000 total author impressions on that day. While the account did not explicitly praise the attacks, most of the tweets on the chain conveyed nationalist overtones, emphasizing the West’s comparatively muted response to PKK attacks in Turkey and to the Assad regime’s brutality in Syria. Calling on Belgium to speak up after the attacks, “we didn’t hear you when bombs were

Islamic State Networks in Turkey

Page 24

going off in Syria,” @Latormenta wrote. “[N]ow that they’ve detonated over there, have you any pain?”

These themes were not limited to the immediate aftermath of IS attacks; they were periodically invoked to raise support for jihadists in Syria. On April 25, for example, the Netherlands-based account of a suspected IS sympathizer, @yasinyener4, tweeted scathing remarks about an alleged PKK regional leader, whom another Turkish account had accused of insulting an Islamic prayer in a shared video. @yasinyener4 referred to the old man as an “infidel dog,” commenting that he would soon pass away and “meet that blessed maker of yours whom you have mocked.”

Similarly, many accounts invoked religious duty to animate their followers. Accounts such as the now-defunct @Abuahmed0666 and @Kurdimuslim36 were recorded tweeting particular Quran verses that stress the importance of jihad for salvation. These tweets either accompanied or followed tweets regarding battlefield developments in Syria by the same authors, and appeared to serve as a tool to recruit militants for the jihad. Of all the tweets captured in our study, these were among the most influential, recording some of the highest number of shares (retweets) or endorsements (likes).

While the study concluded its data collection in May 2016, FDD closely watched Twitter following the Reina nightclub attack on New Year’s Eve 2017. Most Turkish social media users condemned the attack, but some justified it with Islamist rhetoric against alcohol consumption and revealing dress for women – sins they accused the Reina socialites of committing on the non-Islamic holiday of New Year’s Eve.143

Following the New Year’s IS attack in Istanbul, Deputy Prime Minister Numan Kurtulmus announced that

143. Ezgi Basaran, “Secular citizens of Turkey have never felt so alone,” The Washington Post, January 3, 2017. (https://www.washingtonpost.com/posteverything/wp/2017/01/03/secular-citizens-of-turkey-have-never-felt-so-alone/?utm_term=.eafc14818b70)144. “Probe launched into 347 social media accounts after Istanbul attack,” Hurriyet Daily News (Turkey), January 2, 2017. (http://www.hürriyetdailynews.com/probe-launched-into-347-social-media-accounts-after-istanbul-attack.aspx?pageID=238&nid=108024)145. Tulay Karadeniz, Ece Toksabay, Orhan Coskun, Tuvan Gumrukcu, and Nick Tattersall, “Turkey questions U.S.-led coalition presence at Incirlik air base,” Reuters, January 5, 2015. (http://www.reuters.com/article/us-mideast-crisis-syria-turkey-usa-idUSKBN14P0XR)

investigations were being launched into 347 social media accounts, and the Union of Turkish Bar Associations said it was preparing to file additional complaints itself.144 Turkey’s jihadists, however, are unlikely to be troubled by such threats. Ankara’s crackdown on social media primarily targets pro-Kurdish activists and government critics. The Islamic State can be expected to continue to exploit this platform, thanks to Ankara’s selective approach and to the group’s simple ability to recreate new usernames when existing ones are expunged.

Security in Post-Coup Turkey: Grim Prospects

On the night of July 15, 2016, rogue elements within the Turkish military rolled their tanks into the streets of Istanbul and began flying their jets over Ankara in an attempted military coup against the AKP government. The perpetrators failed to capture Erdogan, who called on the people to take to the streets and resist the coup. Between the president’s call, the mosques, and social media, thousands of Turks mobilized. They flooded the streets and reclaimed public areas held by the mutineers. The coup effort failed within hours, but it continues to have major implications for Turkey.

Erdogan and the AKP promptly blamed the plot on the followers of Fethullah Gulen, a U.S.-based Turkish cleric. Ankara demanded that Washington immediately extradite the cleric to Turkey, and pro-government dailies launched a media campaign alleging direct CIA complicity in the attempted coup. In addition to skyrocketing anti-Americanism in Turkey, Ankara’s efforts to coerce the U.S. into extraditing Gulen – including by challenging the legitimacy of the U.S.-led coalition’s presence at the Incirlik airbase145 – has severely strained the country’s ties with its top NATO ally.

Islamic State Networks in Turkey

Page 25

Meanwhile, the AKP unleashed a massive purge to clear Gulenists from the Turkish bureaucracy. More than 100,000 civil servants were either sacked or detained within weeks of the coup attempt. Having declared a “state of emergency,” the president issued decree after decree shutting down businesses, media organizations, and charities. The political scene in Turkey now increasingly resembles a witch hunt, with the government targeting opposition figures with no ties to the Gulen network, including Kurdish activists, journalists, novelists, and academics. The purges have drained human capital from Turkey’s vital institutions, notably the education ministry and the judiciary.146

But the worst damage may have been done to the country’s security and intelligence services, at a time Turkey needs them most. On July 28, only two weeks after the coup attempt, U.S. Director of National Intelligence James Clapper warned that Turkey’s purges of the military were “harming the fight against Islamic State,” referring mostly to the disruption of U.S. operations from the Incirlik airbase as well as the purge of several key Turkish air force members working closely with NATO.147

The Turkish Armed Services, currently at war with Kurdish militants in the Turkey’s east and with both Kurdish and Islamic State fighters inside Syria, has been hit the hardest. Within two months of the coup attempt, nearly 40 percent of all Turkish generals, mostly brigadier generals and rear admirals, had been removed from their posts and replaced with less experienced colonels. The Turkish air force lost nearly half of its combat pilots, reducing the average number

146. “Purge in Numbers,” Turkey Purge, accessed February 2, 2017. (http://turkeypurge.com/purge-in-numbers)147. Warren Strobel, “Turkey military purge harming fight against Islamic State: Clapper,” Reuters, July 28, 2016. (http://www.reuters.com/article/us-turkey-security-usa-purge-idUSKCN1082U8)148. Metin Gurcan, “How post-coup purges depleted Turkey’s military,” Al Monitor, September 16, 2016. (http://www.al-monitor.com/pulse/originals/2016/09/turkey-military-needs-two-year-fill-ranks-emptied-by-purge.html)149. “Purge in Numbers,” Turkey Purge, accessed February 2, 2017. (http://turkeypurge.com/purge-in-numbers); “Turkey removes 1,218 gendarmerie from duty: interior ministry official,” Reuters, November 3, 2016. (http://www.reuters.com/article/us-turkey-security-gendarmerie-idUSKBN12Y0JA)150. “Purge in Numbers,” Turkey Purge, accessed February 2, 2017. (http://turkeypurge.com/purge-in-numbers)151. Josh Keller, Iaryna Mykhyalyshyn, and Safak Timur, “The Scale of Turkey’s Purge Is Nearly Unprecedented,” The New York Times, August 2, 2016. (http://www.nytimes.com/interactive/2016/08/02/world/europe/turkey-purge-erdoğan-scale.html?_r=0)152. Humeyra Pamuk, “Turkey dismisses 87 members of national intelligence agency: media,” Reuters, September 27, 2016. (http://www.reuters.com/article/us-turkey-security-intelligence-idUSKCN11X0TH)

of pilots per warplane from two to less than one. A prominent former Turkish military adviser commented in September that government’s response to the coup attempt has “depleted” the armed services, forecasting at least a two-year period for recovery.148

In total, nearly 8,000 military officers have been dismissed since July 15, and according to the Interior Ministry, more than 1,200 personnel from the gendarmerie have been removed from their posts.149 The Turkish police force, meanwhile, has lost more than 21,000 personnel to the purges.150 With the total number of police officers standing at a mere 9,000, The New York Times likened it to “firing every police officer in Philadelphia, Dallas, Detroit, Boston and Baltimore.”151 In addition, 87 members of the national intelligence agency have been dismissed.152

This instability and polarization in Turkey will likely serve to benefit IS. The group has already been exploiting the country’s ethnic-nationalist divides to recruit both Turks and Kurds to combat the PKK and its affiliates. The Reina attack on January 1 also exposed a secular-religious fault line: Government incitement against New Year’s celebrations, which Islamist Turks denounce as a Christian practice, may have inspired the attack.

Turkey’s military efforts to combat the Islamic State in Syria are also raising the stakes. Clashes in al-Bab are likely to generate revenge attacks against Turkish citizens at home, like the attack at Reina. In early February 2017, Turkish police caught four IS militants in Gaziantep in possession of two dozen suicide bomber belts and 14

Islamic State Networks in Turkey

Page 26

kilograms of TNT, along with Kalashnikov rifles and other materials.153 With threats from inside Turkey, a region in chaos, and a run-down security apparatus since the post-coup purges, Ankara’s ability to thwart major attacks is increasingly in doubt.

RecommendationsThe Turkish government has not prioritized its fight against the Islamic State. Ankara’s failure to effectively fight IS at home and abroad has not only increased Turkey’s vulnerability to terror attacks, but also strained its relationship with the United States. The following eight recommendations are steps the U.S. government can take to help Ankara curtail the jihadist threat to Turkey’s security and stability, and strengthen cooperation between the two NATO allies against the Islamic State.• Given that the Turkish-PKK conflict remains the

top obstacle for Turkey’s full cooperation in the fight against IS, the United States should leverage its relationship with Kurdish groups to urge a ceasefire between the PKK and Turkey and an eventual relaunch of the peace process. The de-escalation of the Kurdish conflict could allow Ankara to return to its pragmatic approach to the PYD, and help Turkey focus its counterterrorism efforts on the fight against the Islamic State.

• In the immediate term, the U.S. should provide support to Turkey in its fight against IS in Syria, but only so long as that mission is confined to fighting IS. Similarly, Washington should provide assistance to its Syrian partners fighting IS so long as that assistance is not used to target Turkey.

• To support Turkey’s anti-IS operations at home, which have improved in the last year, the U.S. should offer Ankara more intelligence on IS suspects, recruiters, financiers, and support networks inside Turkey. Washington should complement this with aggressive diplomacy to ensure Ankara’s commitment to

153. “Police detain four ISIL suspects, foil terrorist attack in Turkey’s Gaziantep,” Hurriyet Daily News (Turkey), February 9, 2017. (http://www.hürriyetdailynews.com/police-detain-four-isil-suspects-foil-terrorist-attack-in-turkeys-gaziantep.aspx?pageID=238&nID=109525&NewsCatID=509)

disrupting all of the Islamist networks that advocate on behalf of IS in Turkey. U.S. diplomats should remind Ankara of the fact that this is in Turkey’s best interest.

• Washington should issue targeted sanctions against all actors in Turkey already identified (including those in this report) as IS operatives, recruiters, financers, advocates, or facilitators, and urge Turkish officials to identify additional financial targets. This is Turkey’s responsibility as a NATO ally and partner in the fight against IS.

• Given the rising level of IS activity in Turkey and Ankara’s lackluster approach to the threat, Washington should consider increasing funding for programs focused on combatting jihadist ideology and de-radicalization in Turkey.

• The United States should similarly expand its efforts to counter IS’s messaging on social media in Turkish. U.S. efforts are primarily focused now on English and Arabic languages. The goal must be to prevent radicalization and recruitment among Turkish speakers.

• To prevent the flow of jihadists and Kurdish militants between Turkey and Syria, the U.S. should offer Turkey technical and material assistance to help improve security along the Syrian border. This assistance could be similar to border security packages provided to Jordan in recent years.

• If Turkey continues to pursue its policies that empower jihadists, Washington must convey that it is prepared to downgrade its relationship with Ankara, including intelligence, military, and perhaps even economic ties. U.S. diplomats should make it clear to their Turkish counterparts that Ankara’s current policies are unsustainable. They endanger first and foremost Turkey itself, as well as broader U.S. interests in Turkey and beyond.

Islamic State Networks in Turkey

Page 27

Appendix ITable I: Islamic State-linked Attacks in Turkey since 2014

Date Location Type of attack Perpetrators Killed WoundedMarch 2014 Nigde Shooting Cendrim Ramadani,

Benyamin Xu,Muhammed Zakiri &unidentified foreign national

3 0

January 2015 Istanbul, Sultanahmet Suicide bombing Diana Ramazova 1 (herself ) 2

June 2015 Diyarbakir, HDP rally Bombing Orhan Gonder 5 (4+1) 400+

July 2015 Suruc Suicide bombing Abdurrahman Alagoz 34 100+

October 2015 Ankara, Peace rally Twin suicide bombings

Yunus Emre Alagoz &Omer Deniz Dundar

102 500+

January 2016 Istanbul, Sultanahmet Suicide bombing Nabil Fadli 10 11

March 2016 Istanbul, Taksim Suicide bombing Mehmet Oztuk 5 36

June 2016 Istanbul, Ataturk Airport Twin suicide bombings and shooting

Rakim Bulgarov,Vadim Osmanov& unidentified foreign national

48 (45+3) 230+

August 2016 Gaziantep, Kurdish wedding

Suicide bombing Unidentified Turkish national

57 90+

January 2017 Istanbul, Reina nightclub Mass shooting Abdulkadir Masharipov 39 70

2014: 1 attack, 3 deaths, no one injured.

2015: 4 attacks, 142 deaths, more than 1,000 injured.

2016: 4 attacks, 120 deaths, more than 350 injured.

2017: 1 attack, 39 deaths, 70 injured.

Total: 10 attacks (8 of them mass-scale), 304 killed, and more than 1,000 injured. Of the 16 perpetrators, 10 were foreign nationals and 6 were Turkish citizens.

Islamic State Networks in Turkey

Page 28

Appendix IITable II: Police Operations Against Islamic State Suspects in Turkey in 2016

* Materials seized (MS), include weapons, ammunition, suicide vests, documents, or digital files.

Date Location Detained Arrested Released Deported DetailsJanuary 6 Yalova, Bursa & Kocaeli 6 TurksJanuary 10 Istanbul 10 TurksJanuary 11 Bursa 12 TurksJanuary 12 Ankara 16 1 Turk 15 SyriansJanuary 12 Adana 9 (4 Turks)January 12 Mersin 9 Turks 3 Turks 6 TurksJanuary 12 Kilis 4 Syrians 2 Syrians 2 SyriansJanuary 12 Sanliurfa 21January 13 Batman 2 TurksJanuary 13 Antalya 3 RussiansJanuary 13 Konya 4 TurksJanuary 13 Izmir 6January 14 Balikesir 5 Turks MSJanuary 14 Osmaniye 1 TurkJanuary 15 Erzurum 3 Turks 2 TurksJanuary 16 Konya 10 TurksJanuary 17 Antalya, Adana &

Osmaniye8 (7 Turks, 1 Syrian)

7 1 Turk

January 20 Kahramanmaras 3January 22 Afyon 6 Syrians 3 Syrians 3 SyriansJanuary 25 Adana 4 foreigners

(3 Indonesian, 1 Saudi)

January 29 Giresun & Rize 2 Turks 2 Turks MSFebruary 10 Adiyaman 4 Libyans 4 LibyansFebruary 13 Adiyaman 1 Turk 1 TurkFebruary 23 Gaziantep 5 Turks 5 TurksFebruary 26 Kilis 1March 3 Bingol 10 Turks 2 Turks 8 TurksMarch 5 Kocaeli 3 Turks 2 TurksMarch 7 Kilis 4 (1 Turk) 3 Syrians 1 TurkMarch 10 Gaziantep 13 7 6March 17 Sanliurfa 1 Syrian 1 SyrianMarch 24 Malatya 4 2

Islamic State Networks in Turkey

Page 29

Date Location Detained Arrested Released Deported DetailsApril 5 Gaziantep 4 Turks 4 TurksApril 22 Adana 2 2April 25 Cankiri 6 Turks 6 Turks MSApril 26 Kocaeli, Kayseri &

Ankara5 Turks 5 Turks

April 30 Antalya 7 foreigners (2 Russians, 4 Syrians)

4 3

May 12 Sakarya 2 Iraqis 2 Iraqis MSMay 13 Batman 3May 16 Malatya 9May 18 Elazig 7 Turks 2 Turks 5 Turks Included a

high-level memberMay 19 Gaziantep Suicide vest

detonatedMay 24 Gaziantep 10 foreigners 10May 31 Antalya 1 Turk 1 TurkJune 4 Bursa 5 Turks 4 Turks 1 TurkJune 7 Yozgat, Ankara &

Istanbul6

June 12 Duzce 5 (2 Turks) 5 June 14 Kahramanmaras 4 (3 Uzbeks) 1 TurkJune 30 Izmir 9 MSJune 30 Kilis 5 (4 Turks) 4 TurksJuly 12 Sanliurfa 4 Syrians 4 Syrians MSJuly 13 Canakkale 6 MSJuly 14 Bursa 1 Turk 1 TurkAugust 6 Adana 16 MSAugust 7 Adana 20August 8 Tekirdag 1 Turk Previously traveled

to Iraq, SyriaAugust 15 Gaziantep 3 3 MSAugust 15 Adiyaman 3 Turks 2 Turks 1 TurkAugust 19 Kilis 9 Turks 6 Turks 3 TurksAugust 22 Mugla 1 Turk MSAugust 25 Nevsehir 3 SyriansAugust 25 Samsun 16 (9 Turks) MSAugust 26 Konya 20 MSAugust 26 Adana 20 August 26 Yalova 5 Turks 2 Turks MS

Islamic State Networks in Turkey

Page 30

Date Location Detained Arrested Released Deported DetailsAugust 26 Ankara 17 16 MSAugust 27 Marmaris 1 Syrian 1 SyrianAugust 29 Elazig 10 Turks 10 TurksSeptember 2 Gaziantep 24 Turks 8 Turks 16 TurksSeptember 5 Ankara 20 16 4 MSSeptember 7 Izmir 9 Turks 7 Turks 2 TurksSeptember 8 Hatay 10 foreigners

(Azeri & Egyptian)

Arrests including IS's Gaziantep cell leader

September 10 Bingol 11 7 MSSeptember 12 Samsun 4 IraqisSeptember 16 Istanbul 24 foreign

nationalsMS

September 18 Istanbul 40 foreign nationals

MS

September 28 Yalova 5September 28 Sinop 3 2October 3 Gaziantep 4October 7 Antalya 8 8 MSOctober 9 Kilis 3 (2 Turks) 2 1October 9 Gaziantep 32 (20 Turks) 14 7 Suicide bomber

detonated bomb, killed 4 police

October 14 Nigde 1 Turk MSOctober 17 Yozgat 1 SyrianOctober 19 Ankara Turkish IS memberOctober 19 Diyarbakir 20October 24 Kastamonu 5 Syrians Linked to

Gaziantep cellOctober 27 Gaziantep, Sanliurfa

& Aydin25 (8 Turks) 23 MS

October 27 Istanbul, Gaziantep, Sanliurfa, Erzurum, Kirklareli & Edirne

81 (21 Turk) 31 locations, MS, 20 of the 60 foreigners who were detained were underage

October 28 Samsun 21 Iraqis MSOctober 28 Eskisehir 1 Turk 1 Turk MSOctober 28 Ankara 1 Turk 1 Turk MS

Islamic State Networks in Turkey

Page 31

Date Location Detained Arrested Released Deported DetailsOctober 28 Malatya 1 IS bomber was

detained

November 2 Bilecik MSNovember 3 Kocaeli 8 SyriansNovember 7 Bitlis 4 Turks 3 Turks 1 TurkNovember 8 Kirklareli 6 foreigners

(Russian, Ukranian, Uzbek)

MS

November 9 Aksaray 2 TurksNovember 10 Sivas 10 Turks 5 Turks MSNovember 11 Nigde 4November 11 Kocaeli 2November 14 Kayseri 12 Turks 3 TurksNovember 15 Corum 1 Iraqi 1 Iraqi High-level member

of ISNovember 15 Tekirdag 3 (2 Turks) 2 Turks 1

foreignerNovember 17 Sakarya 18 (Iraqis/

Syrians)MS

November 23 Denizli 1 Syrian 1 SyrianNovember 25 Mersin 3 (1 Turk) 3November 27 Istanbul 11 foreigners MSNovember 28 Samsun 8 IraqisDecember 2 Sakarya 3 Iraqis 3 Iraqis MSDecember 2 Mugla 1 SyrianDecember 3 Agri 1 Russian 1 RussianDecember 13 Tokat, Istanbul,

Bursa, Ankara/ Gaziantep & Sanliurfa

9

December 13 Van 12December 16 Konya 43 Turks 43 TurksDecember 24 Kilis 7 foreigners 7 foreignersDecember 28 Adiyaman 3 TurksDecember 29 Sakarya 7 Iraqis 2 assassins of IS,

1 cyber of IS,4 fundraisers of IS

December 30 Gaziantep 18 MSDecember 31 Ankara 8 MS

Islamic State Networks in Turkey

Page 32

Total: 54 ProvincesAdanaAdiyamanAfyonAgriAksarayAnkaraAntalyaAydinBalikesirBatmanBilecikBingolBitlisBursaCanakkaleCankiriCorumDenizli

DiyarbakirDuzceEdirneElazigErzurumEskisehirGaziantepGiresunHatayIstanbulIzmirKahramanmarasKastamonuKayseriKilisKirklareliKocaeliKonya

MalatyaMarmarisMersinMuglaNevsehir NigdeOsmaniyeRizeSakaryaSamsunSanliurfaSinopSivasTekirdagTokatVanYalovaYozgat

Map of Table II: Police Operations Against Islamic State Suspects in Turkey in 2016

Anti-IS operations

KEY

No anti-IS operations reported

Islamic State Networks in Turkey

Page 33

Appendix III: SGA Turkish Social Media StudyMain Findings and Executive Summary

The following report discusses the main findings from data collected by Sidar Global Advisors through Brandwatch on the Twitter presence of Turkish ISIS operatives/sympathizers between March 15 and May 1, 2016.

An Alarming Development: Introductory Notes & General Findings

Between March 15 and May 1, 2016, our project actively tracked the Twitter accounts of 137 Turkish Islamic State (ISIS) members/sympathizers, along with 18 pre-established keywords, which were Islamic concepts often employed by ISIS, transliterated from Arabic into Turkish—i.e., “Cihad” instead of “Jihad.” In total, our social media analytics technology, Brandwatch, recorded nearly 694,000 mentions from these accounts and keywords. Automated sentiment analysis scored around 20,000 mentions as generally positive, and 83,000 were scored as negative. The main topics or concepts of interest ranged from general terms such as “Allah” and “Islam” to more specific alliterations.

The following were the most discussed issues by mention:1. ISIS & the Islamic State: 55,000 mentions2. Allah: 50,000 mentions3. Islam: 46,000 mentions4. Erdogan: 22,000 mentions5. AKP: 20,000 mentions6. El-Nusra (al-Nusra): 18,000 mentions7. Hilafet Turkiye (Caliphate Turkey):

13,800 mentions

The following queries had the highest number of mentions:1. Cihad (Jihad): 182,000 mentions2. Kafir (Infidel): 130,000 mentions3. Halife & Hilafet (Caliph & Caliphate):

89,000 mentions

4. Nusra : 49,000 mentions5. Kudus (Jerusalem): 48,000 mentions6. Siyonist (Zionist): 37,000 mentions7. Mucahid (Mujahid): 34,000 mentions8. İslam Devleti (Islamic State): 30,000 mentions

1. Data Collection

This study tracked 18 Turkish queries or keywords on Twitter and collected the tweets wherein they appeared. FDD developed the queries based on Turkish concepts used frequently by ISIS members and sympathizers. Where necessary, different spellings of the keywords were taken into account, e.g. “Mucahid” and “Mucahit.” The study did not, however, check for tweets with spelling errors.

In addition to keyword collection, SGA also tracked the accounts of 137 ISIS sympathizers tweeting in Turkish. We used non-probability snowball sampling to identify the accounts of these 137 sympathizers. First, SGA’s Turkish analysts worked with FDD to identify a small sample of Turkish-language Twitter accounts that were clearly sympathetic to ISIS based on their profiles and tweets. From these individuals, we expanded our sample to other Turkish-language ISIS sympathizers in their Twitter networks.

FDD commissioned this study to begin data collection on March 19, when a Turkish suicide bomber with ISIS links struck a busy tourist area in central Istanbul killing at least four people. The initial plan was to collect data for one month, but we extended the data collection process for another two weeks to expand the dataset, and ended the data collection on May 1, 2016.

SGA used its technology partner company’s leading social media monitoring and analytics platform, Brandwatch Analytics, to collect data for the study. The platform systematically crawls the internet to create a database of social media accounts and websites.

Islamic State Networks in Turkey

Page 34

We specifically limited our data collection to Twitter upon FDD’s request. Based on our initial queries and 137 specified Twitter accounts, Brandwatch crawled Twitter every 5 to 6 hours. The platform saves the new URLs and accounts in its first-level archive. Metadata are applied to these URLs and accounts (e.g. number of visitors, MozRank, Backlinks, etc.) and then saved in the second-level archive. Queries search the second level of storage to assure that the application works as fast as possible.

In total, we recorded 693,533 mentions from these Twitter accounts and keywords between March 19 and May 1 (each tweet that contains content that matches a query counts as a mention). Of the 137 tracked accounts, most were shut down by either Twitter for having inappropriate imagery or content, or were reported as hazardous and forcibly terminated. However, we were able to retain the data from these accounts for the period from March 15 to May 1, which are still available in our archives.

Tracked Queries:• Ahrar (Ahrar [al-Sham])• Cihad (Jihad)• Halife (Caliph)• Hilafet (Caliphate)• İslam Devleti (Islamic State) • İstişhadi (Prosperity)• Kafir (Infidel)• Kefere (Infidels)• Konstanniye (Istanbul)• Kudüs (Jerusalem)• Mücahid (Mujahid) • Mürted (Apostate)• Müşrik (Polytheist)• Nusra ([Jabhat al-] Nusra)• Rafizi • Siyonist (Zionist) • Tağut (Tyrant)• Tekfir (Takfir)

2. Data analysis

We used six primary types of analysis to answer the research questions for this study. Much of the data analysis was conducted using Brandwatch’s analysis tools. SGA Turkey analysts conducted additional qualitative analysis to describe how the data fit within the broader landscape with the Turkish Twitter universe.

Volume Analysis

Volume analysis is the most basic inquiry that we conducted for this study. Volume analysis provides counts and descriptive statistics on the number of times and proportions that a queried keyword is mentioned on Twitter during the date range of our study. The volume can be analyzed by a specific date, author, topic, and sentiment.

Topic Analysis

We also conducted topics analysis. This type of inquiry identifies up to 50 top terms or phrases most associated with an author (account) or query (keyword) within our study’s date range. The topics are based on a random sample of up to 600 query mentions but do not include the query itself. Topics are calculated in two ways. The first is the overall volume of mentions. The second is burst: This is calculated by dividing your selected date range into two equal parts and comparing the increase in volume of that topic across the two halves. E.g. if there were 50 mentions of a topic in the first half, and 75 in the second, that is an increase of 50% and therefore the topic would have a burst value of 50. Topic analysis can be displayed as a word cloud that highlights the 30 most frequently-used words or phrases within the study’s the date range. It can also be displayed in a table. Sentiment analysis can be paired with both table and word cloud topic analysis to show the positive or negative value of the most frequently-used words or phrases.

Islamic State Networks in Turkey

Page 35

Sentiment Analysis

Our technology supports sentiment analysis in 28 languages, including Turkish. The platform automatically scores all mentions as “positive,” “negative,” or “neutral” based on a combination of manual and automated natural language processing (NLP) techniques. The system scores sentiment conservatively, only classifying mentions that it is reasonably certain are positive or negative. Mentions for which the sentiment cannot be classified with enough accuracy are left as “neutral” (or “unclassified”). Therefore, most of the mentions in our project were considered “neutral.”

Author Analysis

Author analysis identifies the most active authors in the data. Within this type of inquiry, we can view all query mentions by author (Twitter account). We can also determine how impactful a specific Twitter account is by its impact score.

Impact is a score that reflects how much a mention, author, or site is seen and shared online. It is a quick way to filter the important from the unimportant. The impact score is comprised of:

Factor A: How much potential a mention has to be seen.

Factor B: How much a mention has been viewed, shared, or retweeted.

To calculate the impact score for an author, we sum up the number of mentions the author wrote and the maximum values achieved for Factors A and B (as outlined above) and then normalize it.

SGA calculated impact scores using metrics such as Kred score, MozRank, and retweets. From this data, we assign an author a score from 1-100. The tweets of an author with a score of 100 are widely viewed and shared.

Geolocation

We also analyzed the location of Twitter accounts with geolocation analysis. We use a 5-tiered scale to locate

authors. First, the platform looks for geo-coordinates provided by the GPS-enabled devices from which an author tweeted. Approximately 1-5% of tweets are tagged with coordinates. This is the most accurate type of geolocation, but this method is not 100% accurate as users can manually change their location. If no GPS coordinates are available, we scan tweets for mentions of a location (e.g. if an author tweets “I made it to Raqqa,” and then tags himself in Raqqa, the platform will geo-locate the author in Raqqa, Syria). Third, location can be analyzed via time zone reporting on user profiles. While this method cannot determine if an author is tweeting from Raqqa or Istanbul, since they are in the same time zone, it can establish if tweets originate in Europe or Asia. SGA can also use web domain suffixes to geo-locate authors. For example, a user posting links with .tr domain suffix might indicate that the author is in Turkey. Finally, if no other information is available Brandwatch will identify users by the IP-address of the web host. This method can be problematic since many web hosts are in the United States. Therefore, our analysts have chosen to filter out all such tweets in order to accurately obtain data from the disclosed samples.

Gender Identification

We also attempted to identify the gender of authors. Our technology determines user gender by first matching names to a dictionary of over 40,000 names, which are then used to assign a gender. If a user’s name is not gender-specific, the platform will not assign a gender.

3. Data Overview And Labeling

Over the course of the 45 days, we gathered a net total of 693,533 mentions from the 18 queries and 137 authors. Unlike states with comprehensive internet censorship policies, the Republic of Turkey’s censorship laws are retroactive—accounts are censored only after Turkish authorities or Twitter users request the termination of these accounts from social media bodies. Our technology allowed SGA to retain censored tweets from cancelled accounts except for the images and link shares.

Islamic State Networks in Turkey

Page 36

To make use of this vast and unprocessed data, this report sought to assess these mentions within their proper general and specific contexts. This allowed our analysts to use filters to deduce any general findings. While determining the general tendencies of the Turkish Twitter universe was not possible given Brandwatch’s limitations, certain general inferences could be made regarding the tendencies of Turkish social media presence.

This report attempted to differentiate the authors into categories. While these terms can be relative, some commonalities between several authors allowed us to attempt a classification.• Islamist: Possesses a theological disposition and

tweets religious content and occasionally political content. While some of these tweets were of interest (i.e. expressed sympathies for some of the grievances in Syria), their utility is of dubious importance for the sake of this report.

û E.g.: “You cannot say that laicite is an imperialist ideology and we are still debating an ideology that states Islam is a back-ward religion!” (sozunhaysiyeti, 28 April)

• ISIS: Either directly affiliated with or sympathizer of ISIS “beyond-reasonable doubt.” Their tweets are of a much more violent and ideological nature than those of the Islamists. In order to gain more insight into this group, this report further divided this category into two sub-components.

Ȋ Violent: Mostly shares violent thoughts or advocates for violence; tweets often contain violent images, videos, and messages.

û E.g.: “We could do ten-fold of what we did at NATO and the heart of EU in Brussels here in Turkey...” (EbuMelhame25, 23 March)

Ȋ Non-Violent (propaganda): Expresses or advocates on behalf of the ISIS; tracks developments in Syria, Turkey, or Europe. Generally refrains from transgressions that would warrant the termination of their account.

û E.g.: “#ISIS #CaliphateTR #KerkukVilayet Visuals from the battle for Allas.” (TevhidMedya, May 1)

4. Discussion

1. The Authors / Accounts

Top authors based on number of tweets:1. AbuuAhmed8: 1,4452. Mucahidakinci: 1,2313. Kurdiselman36: 1,0594. AliReddi: 9855. DVIslam: 9806. MuhammedMuhajer: 9017. Abu_ahmed0666: 7958. Dogrucu_Mucahit: 7229. Kurdimuslim36: 72010. 72Chd: 678

Regardless of their Twitter mentions, some accounts were more influential and had a greater impact (viewed, re-tweeted, mentioned) than those that merely had a high volume of tweets.

Top influential authors with their impact variable (tweet, tweet audience, retweet, retweet audience, audience of initial audience, etc.):1. Cihatolog: 892. Thesporcu: 883. Cihat_bayram: 874. 72Chd: 875. MuhammedMuhajer: 866. Dogrucu_Mucahit: 857. Nagihansaka: 858. 57Islamicc: 849. Hayallerimvben: 8410. Tevhidmedya_: 84

Of these accounts, most were shut down by either Twitter or were reported as hazardous and forcibly terminated. Brandwatch was able to retain the data for these accounts from March 15 to May 1.

Not all of these accounts had a role to play within the keyword or specific account data. As far as the general

Islamic State Networks in Turkey

Page 37

aggregate data is concerned, their contributions mattered and, in effect, we were able to attain the sum-total of all things worthy of mention during the timeframe of this study. However, since some of these accounts were willfully neglected or simply relegated to irrelevance, our scoped data for the queries and tracked specific Twitter accounts did not necessarily involve their contributions. This way, we were able to apply an active filter of whom to track and why, as opposed to involving all which, directly or indirectly, could have confounded our main findings (e.g. influential accounts using the above terms for satirical purposes were left out).

Some of these Twitter accounts acted as “post-boxes” in which different authors deposited nuggets of information for their audiences for a more coherent online presence. Hence, some accounts are not necessarily accounts of individuals but act as “umbrella accounts” that pool the sources together.

Figure 1: JN_Cendel2’s tweets are examples of many re-direct tweets, which inform about terminated accounts and re-direct followers to the renewed addresses.

Figure 2: Tevhid Medya is an “umbrella” account that uses hashtags to organize a steady stream of information all under one roof. These accounts act as alternatives to formal news agencies.

Our report concluded that of the 21 most impactful authors, 7 (33%) were Islamists, while the remaining

67% could be classified in a broad sense as ISIS sympathizers—of which 9 (43%) were militant and 5 (24%) and were non-militant.

From the 137 authors, a total of 97 (71%) were either shut down or abandoned at one point or another. Our analysts scoured the terminated accounts, and determined that 39 (40%) of those censored were of a militant-ISIS disposition.

2. Topic And Sentiment Analysis

Brandwatch recorded 693,533 mentions from the 137 accounts and 18 keywords. Automated sentiment analysis scored only 3% of the data, with 20,038 mentions, as generally positive. Another 12%, with 82,856 of the overall mentions, were scored with negative sentiment. The remaining 85% of the data, or 590,639 mentions, were of neutral value.

Of the general data, Brandwatch recorded the most mentioned topics as “Allah” (49,703 mentions), “Islam” (46,235 mentions), “ISID” (24,273 mentions), “Turkiye” (24,273 mentions), and “Halep” (23,117 mentions). The least mentioned topics were “Davutoglu” (2,805 mentions), “Laiklik” (4,803 mentions), “Hz” (5,779 mentions), “Kuzey Halep” (6,935 mentions), and “IslamDevleti Hilafet TR” (6,935 mentions). Sentiments for these topics could not be obtained.

Our search recorded a total of 694, 779 tweets containing one or more of the 18 queries (or keywords). The average mention per keyword came out to 38,598. Taking this average into consideration, this report found the three queries most mentioned above average to be: “Cihad” (182,161 mentions), “Kafir” (130,041 mentions), and “Nusra” (48,867 mentions). According to these numbers, the word “Cihad” was mentioned in more than a quarter (26%) of our total sample of tweets.

Of the queries tracked, the following had the highest negative mentions: “Kafir” (Infidel) had 23,038 negative mentions, “Cihad” (Jihad) had 15,765, “Nusra” had 8,305, and “Siyonist” had 7,615. The most negatively mentioned word in relation to total mentions was the

Islamic State Networks in Turkey

Page 38

query for “Siyonist” (21%). The average negative value for all queries was 8.6%.

Queries with the highest positive mentions were: “Cihad” (6,482 mentions), “Kafir” (3,576 mentions), “Kudus” (1,764 mentions), and “Mucahid” (1,599 mentions). The most positively mentioned word in relation to total mentions was the query for “Mucahid” (nearly 5%). The average positive value for all queries was 3.75%.

In addition to the figures above, our search recorded a considerable number of tweets discussing the Kurds or the Kurdistan Workers’ Party (PKK), a designated terrorist group that has been fighting an insurgency in Turkey for decades. The tweets by ISIS or ISIS sympathizer accounts often incited against the Kurds as well as against the West, which is seen by Islamists and nationalist Turks as a patron of the PKK. They also appeared to use Turkish grievances against the PKK to galvanize support for jihad among their Turkish-speaking audience.

Figure 3: Tweet from an Islamist-Turkish Dutch account, @1ali2demir2, after the Brussels airport attack: “You should have protected your airport instead of the atheist PKK terror organization’s tent.”

While some tweets targeted the Kurds, most recorded tweets were anti-PKK. The authors often labeled the group and its supporters as infidels, unbelievers, or atheists. The PKK has its roots in Marxist-Leninist ideology and is thus often associated with atheism. However, the authors’ emphasis on the un-Islamic character of the group is particular to Islamist rhetoric.

Figure 4: @Yasinyener4 of the Netherlands refers to a PKK official’s mockery of Islam by using terms as “infidel,” “dog,” and states that “…you will soon meet your maker and face justice.”

In many instances, anti-Western Turkish sentiments came hand-in-hand with anti-PKK rhetoric, or came in the form of accusations against the Europeans for their hypocrisy in supporting the PKK while demanding that Turkey do more to combat ISIS. The tweets suggest an underlying view that Turkey and the West have two distinct enemies—the PKK for the former, ISIS for the latter—and they downplay the ISIS threat to Turkey. The patterns reflect a binary worldview, in which Europe, the PKK, and the Kurds are presented as the other, suggesting nationalist undertones.

These kinds of associations allow pro-ISIS or sympathizer accounts to recruit among Turkish ultranationalists by presenting the jihadists as crusaders against the infidel

Islamic State Networks in Turkey

Page 39

enemies, both the West and the PKK alike. The West is implied to deserve the attacks due to its silence on, and thereby complicity in, the PKK’s attacks against Turkey, as well as the Assad regime’s attacks on Syrian rebels and civilians.

Figure 5: Highly influential Latormenta (now defunct) tweeting about the Brussels attacks on March 22 (total impressions of author on March 22: 99,702): “Hello Belgium, raise your voice. We didn’t hear you when bombs were going off in Syria.” Most tweets have nationalist overtones, emphasizing the West’s muted response to PKK attacks in Turkey or of the Assad regime in Syria.

3.Geographic Distribution

Geo-tagged location disclosure findings were limited as Brandwatch could only register 1% of the data with geo-tagged location coordinates. Hence, the ceiling on geo-tags was capped at 5,000. Of the geo-tagged mentions associated with the queries, our report found that an average of 85% of these were within the borders of Turkey. The remaining 15% came from around the world. Most of the mentions recorded outside of Turkey registered from within the Belgium-Netherlands-Germany (BNG) region. This cluster in the general data segment recorded 291 mentions.

The three most geo-tagged queries within Turkey were “Kefere” (93%), “Ahrar” (92%), and “Mucahid” (92%). The most query mentions outside Turkey were “Nusra” (30%), “Musrik” (28%), and “Siyonist” (27%). The least mentioned geo-tagged queries within Turkey corresponded with the most geo-tagged mentions from abroad.

The vast amount of geo-tagged mentions in Turkey occurred in the urban regions of Western Turkey—Istanbul (2,527 mentions), Izmir (420 mentions), and Ankara (388 mentions)—along with some cities in Eastern Turkey registering high numbers of mention in relation to populations, such as Adana (243 mentions).

A random sampling of 5 geo-tagged locations of queries allowed our report to further differentiate between East and West Turkey: an average distribution ratio of 20% (Eastern Turkey) and 80% (Western Turkey) was observed within the country.

Figure 6: Region 1, the Marmara region, always registers a high throughput with pro-ISIS mentions, as do Regions 2 (Ankara region) and 5 (the Turkish cities on the Syrian border, Gaziantep and Sanliurfa). The Black Sea Trabzon region (3 on the map), only occasionally lights up with Islamist mentions.

Based on the sampling, the words “Siyonist” (24%), “Kafir” (22%), and “Mucahid” (22%) clocked the highest mention percentages from Eastern Turkey, whereas “Kudus” (86%), “Kafir” (83%) and “Cihad” (80%) recorded the highest rates of mentions in the west.

Figure 7: Distribution of 1,322 registered mentions of “Kafir” (Infidel-Heretic) in Turkey.

Islamic State Networks in Turkey

Page 40

Figure 8: Distribution of 506 registered mentions of “Kudus” (Jerusalem) in Turkey.

Figure 9: Distribution of 460 registered mentions of “Halife” (Caliph) in Turkey.

Geo-tagged mentions from abroad from this sample did not yield a statistically consistent pattern but were interesting nonetheless. The BNG area recorded an average of 31% from all mentions abroad, making it an area with a relatively high mention-to-km ratio. From within this cluster, the mentions of “Siyonist” (59%) and “Kafir” (50%) were highest. The least recorded mention from within this area was “Cihad” (12%).

Figure 10: Turkish mentions of the term “Cihad” appeared beyond the expected range in the Ruhrpot (Dusseldorf, Koln, and Frankfurt) region, along with the Dutch Randstaad (Amsterdam, Rotterdam, and Den Haag) region, and Brussels, Belgium.

Figure 11: “Kafir” also appeared in the Netherlands and Germany, and to a lesser extent in Belgium.

4. Trigger Events & Moments

This report observed a correlational relationship between the queries, authors, and global real-time developments in three main general contexts: terror attacks and their aftermath during the period of March 19 to March 23 (the Istanbul (Taksim) and Brussels (airport) suicide bombings), important dignitary visits from states involved in Syria (Iran and Saudi Arabia), and other important tactical/operational developments in Syria.

Whenever such events took place, many of the 18 queries registered abnormal spikes in activity; namely, these bursts of mentions were above the normal mention averages and, after a period of high intensity, often normalized back down to the average or dipped just below the normal levels of mentions.

Figure 12: Mention spike for “Siyonist” (Zionist) from March 19-22. Most of those killed in the Istanbul bombings were Israeli citizens. Most of the mentions for “Siyonist” abated after the attacks.

Islamic State Networks in Turkey

Page 41

For example, the query “IslamDevleti” (ISIS) was mentioned an average of 624 times per day. During the period of 4 days between the terror bombings in Istanbul (on March 19) and in Brussels (on March 22), however, the mention rate rocketed to figures as high as 1,352 per day. Such spikes translated into a 217% jump from the daily average. Double spikes within short periods only further confirm the relationship between the events and reactions to these query mentions.

Figure 13: Mention spike for “IslamDevleti” (ISIS) from March 19-22. These spikes are essential in “claiming” the social sphere by partaking in the discourse.

To make sense of these mention escalations, this report also examined the sentiment values of the spike dates with “IslamDevleti.” The two sentiment values for the spike days registered well above average values. However, each query and mention spike requires an individual analysis with their dates of importance. A general conclusion on all spiking mentions could not be discerned from single or inter-query analyses.

Figure 14: Prolific author Sozunhaysiyeti (still active), criticizing the Turkish government’s moderate reception of the Iranian president at the 13th Organization of Islamic Cooperation (OIC) summit from April 10-15 in Turkey.

Figure 15: Ahde_Vefa’s (still active) concept map of the most used terms, both hashtaged and not, with “Aleppoisburning” having the most mentions (338). This is similar to Solhanli4 (246 mentions of Aleppo), JN_Cendel2 (209 mentions of the situation in Falluja), and MuhammedAtta4 (110 mentions). Domestic developments also garnered specific attention: “Diyarbakir” had 413 mentions from Hur_Alem and 70 from Gazzeli_Ilyas.

Figure 16: Teror_Gercegi is one of many accounts detailing the tactical and operational developments on the ground in Syria. This level of attention to tactical details at the village level in Syria demonstrates how well in-tune these Turkish accounts are with real-time military developments.

5. Recruitment & Resistance To The Islamic State

Even though many pro-ISIS and radical Islamists exploit the social media sphere to their advantage—or, rather, are allowed to depending on how the matter is examined—an active and rather vigilant

Islamic State Networks in Turkey

Page 42

counter-ISIS presence also exists from those that wish to challenge these Islamist authors with their assertions, facts, and arguments.

Figure 17: Dirilis_TR targets “Laiklik” (Secularism).

Of the pooled data, many influential authors such as Cihatolog and Nagihansaka, among others, seemed to directly challenge Islamist or ISIS accounts every step of the way. Our automated data incorporated a multitude of mentions that confounded our data, but yielded an important finding nonetheless: Turks are prolific social media users and the social media realm itself is a front within this war of discourse.

This report discerned many tailored tweets addressed to the Islamist audience at large with the sole purpose of swelling the rank-and-file of ISIS. More often than not, these tweets emphasized the “glorious” sacred cause and virtues of martyrdom. Though direct tweets with proper nouns or out-right statements were lacking—on account of the existing censorship—those authors that were able to call for more “mujahedeen” were quite influential and impactful (e.g., they had many followers that retweeted and/or mentioned those authors directly).

When assessing the Turkish audience these authors are directly pleading to, we include not only native Turks, but also those who are able to understand Turkish and join the cause. In this regard, we observed a panoply of Chechens, Islamist Turkic, and native Turks who either partook in the recruitment, spread the word on social media, or traversed through Turkey to fight in Iraq.

Active calls for direct recruitment were hardly ever found, yet a few notable authors called on their audiences to take action after sharing images of war victims in Syria.

Figure 18: Discussions on possible recruitment by way of “calling for Jihad” and “going to the front” questioned in such tweets as Muhaysini_3.

Figures 19 & 20: Quran verses used as galvanizing recruitment instruments. Here, authors Abuahmed0666 & Kurdimuslim36 mention verses on how those that fight for the cause of Allah are worthy of his affection. These authors refer to these select verses amidst other tweets about the situation in Syria—within this context these allusions gain ominous connotations.

Figure 21: Though not as common, tweets such as this one by Ahde_Vefa_ carry a message of recruitment by way of praising a select group of terrorists in Syria. The use of a poster-child group is common when recruitment for a cause is at hand; militaristic praise is used as a norm-setting device in which would-be recruits are conditioned to want to be more like that trend-setting group.

Islamic State Networks in Turkey

Page 43

AppendixThe following are the 20 most influential Twitter authors with their total owner and audience impressions, dates of intense presence on social media, and their most discussed topics.

1. Sozunhaysiyeti (6,500,000)

• History: April 12-14 April 20-22 April 27-28

• Topics: Islam Allah Erdogan

2. Gazzeli_Ilyas (2,985,000)

• History: April 12-14 April 15-17

• Topics: Demirtas Istanbul Ahlaksizligasessizkalma (dontstaysilenttoimmorality) Diyarbakirdakisapkinligadurde (saystoptotheheresyinDiyarkakir)

3. Yahyakurdi001 (2,600,000)

• History: March 22-25 • Topics: ISID (ISIS) Irak (Iraq) ABD (USA)

4. Mazlum_Direnis (2,300,000)

• History: April 6-9 April 18-23

• Topics: Allah Yerigelmiskensoyleyeyim (Letmejustsay)

5. El_sevap (1,824,000)

• History: March 23-24 March 28-30 April 3-4

• Topics: Suriye (Syria) PKK Rusya (Russia) ISID (ISIS)

Islamic State Networks in Turkey

Page 44

6. Muhtar1334 (1,621,000)

• History: March 31 - April 2 • Topics: Irak (Iraq) ABD (USA) Turkiye (Turkey) Musul (Mosul)

7. Ahde_Vefa_ (1,450,000)

• History: April 12 April 16 April 30

• Topics: Halepyaniyor (Aleppoisburning) Suriye (Syria) Suriyedekatliamvar (ThereisamassacreinSyria)

8. Muhaysini_3 (950,000)

• History: March 26-28 April 1-2

• Topics: AbdullahMuhaysini Seyh (Sheikh) Seyhmuhaysini Cihad (Jihad) Allah

9. Almahamamedia (917,000)

• History: March 18-23 March 26 March 30-31 April 8

• Topics: Rusya (Russia) Suriye (Syria) ISID (ISIS)

10. Latormenta_ (850,000)

• History: March 22-24 • Topics: ISID (ISIS) PKK Belcika (Belgium) ABD (USA) Turkiye (Turkey)

Islamic State Networks in Turkey

Page 45

11. MuhammedAtta4 (833,000)

• History: March 25-28 • Topics: ISID (ISIS) Kaide ([al-] Qaeda) ABD (USA) Esed ([Bashar al-] Assad) Nusra ([al-] Nusra) Felluce (Fallujah)

12. Buruciye5 (720,000)

• History: March 31 - April 1 April 7-8

• Topics: Islam Hilafet (Caliphate) Cihadinyahudileri (theJewsofJihad)

13. JN_Cendel2 (620,000)

• History: March 27-28 March 30-31 April 17

• Topics: Felluceyeacilinsanyardimi (urgenthumanaaidtoFallujah) Kafir (Infidel) Suriye (Syria) ISID (ISIS) PKK

14. Skarsavi (613,000)

• History: April 1-2 • Topics: Seyh (Sheikh) Ummetin (oftheUmma) AKP

15. Terror_gercegi (558,000)

• History: March 28-31 • Topics: Irak (Iraq)

Suriye (Syria) Islam Devleti (Islamic State (ISIS))

Islamic State Networks in Turkey

Page 46

16. Muslumanlaz (555,000)

• History: March 24-26 April 3-5

• Topics: Islam Devleti (Islamic State (ISIS)) Dimesk eyaleti (the Damascus state) Esed ordusu (the Assad army)

17. 13elyesa13 (495,000)

• History: March 27 April 10 & 13

• Topics: ISID (ISIS) PKK TSK (Turk Silahli Kuvvetleri) (Turkish Armed Forces)

18. Solhanli4 (495,000)

• History: March 24-26 April 3-4

• Topics: Halep (Aleppo) Acil (Urgent) IslamDevleti (Islamic State (ISIS)) Askerolduruldu (soldierwaskilled) PKKliolduruldu (PKK[member]killed)

19. Abu_ebu (463,000)

• History: March 22-24 • Topics: Islam Devleti (Islamic State (ISIS)) Suriye (Syria) Irak (Iraq) ABD (USA) Palmyra

20. Hur_alem (424,000)

• History: March 29 April 13

• Topics: Diyarbakirdatarihigun (historicdayinDiyarbakir) Diyarbakirdasapkinligadurde (saystoptotheheresyinDiyarkakir) PKK

Islamic State Networks in Turkey

Page 47

Concepts & General Findings

The following are the top 9 most mentioned queries with any data on usage frequency, location, gender usage, sentiments, and topics.

1. Cihad (Jihad) (182,000)

• Data: Predominantly always high Spike: March 18-24, then falls

Spikes again: March 26-28, then falls to regular intensity• Location: All of Turkey, very active NL - BE - GR

• Topics: Cihat Cagrisi (Call for Jihad) (8,100) Erdogan (7,330) ABD (USA) (7,000) Baris Cagrisi (Call for Peace) (4,500)

2. Kafir (Infidel) (130,000)

• Data: Spike: March 18-22 (corresponds with Istanbul Taksim and Brussels attack) Spikes again April 18-20 corresponds with foreign NATO – Merkel Istanbul/Turkey visits)

• Location: All of Turkey NL - BE - W. Germany

• Mostly men (70+%)• Topics: AKP Erdogan Davutoglu Halep PKK

3. Nusra (49,000)

• Data: Gradual spike: March 18-22 & 26-30 Spikes again: April 2-3 Serious spike: April 10-11, then major slowdown

• Location: Istanbul, Ankara, Syrian border (Antakya)• 72% men• Topics: El-Nusra (35,000) ISID (ISIS) (13,600) Suriye (Syria) (6,500) Turkiye (Turkey) (5,100) Boko Haram (3,500) Halep (Aleppo) (3,600)

• Sentiment: ‘Nusra’ more used by critics of AKP and ISIS

Islamic State Networks in Turkey

Page 48

4. Kudus (Jerusalem) (48,000)

• Data: Spike: March 24-25, then declines for a long time Spikes again: April 13-14, does not decline

Serious spike: April 17-18, then falls Small spike: April 25-27

• Location: All over Turkey Some NL Less GR and BE

• 72% men• Topics: Kuduste (in Jerusalem) (5,500) Aksa ([al-] Aqsa [mosque]) (3,700) Halep (Aleppo) (2,900) Filistin (Palestine) (2,700) Basbakan (Prime Minister) (1,700) Osmanli (Ottoman) (1,198)

• Sentiment: Kudus used as rallying call approximately 8 times.

5. Halife (Caliph) (45,000)

• Data: Spikes: April 1-2 & 13-15 & 24-25• Location: Benelux area All of Turkey London, a bit

• 75% men• Topics: Erdogan (5,000+) Islam The One Allah Sultan Prophet

6. Hilafet (Caliphate) (44,000)

• Data: Overall very low number of mentions Spike: April 18-21 (±700%)

Lower spike: April 24-26• Location: All of Turkey NL - BE - W. Germany

• 73% men• Topics: Islam Devleti (Islamic State)

Hizbut Tahrir Seriat v. Laiklik (& Laikligikazanacagiz) (Sharia v. Secularism)

• Sentiment: Equal, leans towards Islamists

Islamic State Networks in Turkey

Page 49

7. Siyonist (Zionist) (37,000)

• Data: Spike: March 18-22, remains high Major spike: March 28-31, then sharply declines and stays low

• Topics: Israel (4,700) Yahudi (2,500) Palestine (3,500) Fetö (2,207)

• 77% men• Location: All over Turkey, especially 3 main cities• Sentiment: Israel viewed negatively

8. Mucahid (Mujahid) (34,000)

• Data: Spike: March 18-23• Location: Predominantly Turkey Some NL

• 80% men• Topics: Turkiye Suriye AKP

Erdogan Turkmen PKK

• Note: Most mentions of this concept due to names of Twitter accounts containing the term Mucahid.

9. Islam Devleti (Islamic State) (30,000)

• Data: Spike: March 19 & 22 + 24, then falls Spikes again: April 10-11, then falls

Spikes again: April 17-18• Location: Almost entirely Turkey (Istanbul and Antep)• 76% men• Topics: Hilafet TR (4,700) Islam Devleti Hilafet TR (4,100)

• Sentiments: Very Islamist and concepts used in a negative way

Islamic State Networks in Turkey

Acknowledgments

We wish to thank FDD’s Eric Edelman, Aykan Erdemir and John Hannah for their feedback and edits, both substantive and stylistic. We also wish to thank our external readers Aaron Stein, Nicholas Danforth, and Aaron Zelin for taking time out of their busy schedules to provide crucial feedback on our work. We are also grateful to Daniel Ackerman and Erin Blumenthal for the graphics, design, and production of this report. This report and the issues discussed within have benefited greatly from the tremendous thought leadership of the advisors to FDD’s Turkey program.

Islamic State Networks in Turkey

About The Authors

Merve Tahiroglu is a research associate focusing on Turkey. Merve supports the work of FDD scholars with Turkish language research and analysis on Turkey-related matters. Her research focuses on Turkey’s foreign policy, domestic politics, and Ankara’s ties to Tehran. Merve’s personal areas of interest include Turkey’s Syria policy and Islamic extremism in Turkey. Born and raised in Istanbul, Merve earned her B.A in Political Science with a concentration in International Relations from Duke University in 2013. Merve is also a contributor to FDD's Long War Journal.

Jonathan Schanzer is Senior Vice President at FDD. Jonathan is part of the leadership team of FDD’s Center on Sanctions and Illicit Finance, which provides policy and subject matter expertise on the use of financial and economic power to the global policy community. Previously, Jonathan worked as a terrorism finance analyst at the U.S. Department of the Treasury, where he played an integral role in the designation of numerous terrorist financiers. A former research fellow at the Washington Institute for Near East Policy, Jonathan has studied Middle East history in four countries. He has testified before Congress and publishes widely in the American and international media.

About the Foundation for Defense of Democracies

The Foundation for Defense of Democracies (FDD) is a Washington, DC-based non-partisan policy institute focusing on foreign policy and national security.

For more information, please visit www.defenddemocracy.org.

P.O. Box 33249Washington, DC 20033-3249(202) 207-0190www.defenddemocracy.org